Ensemble-Learning-Comparing-Different-Models.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Başlıklar\n",
    "\n",
    "* [Veri Seti](#Veri-Seti)\n",
    "    * [Veri Seti Tanımı](#Veri-Seti-Tanımı)\n",
    "    * [Veri Setine Genel Bakış](#Veri-Setine-Genel-Bakış)\n",
    "    * [Veri Setinin Görselleştirilmesi](#Veri-Setinin-Görselleştirilmesi)\n",
    "    * [Kayıp Değer Analizi](#Kayıp-Değer-Analizi)\n",
    "    * [Veri Ön İşleme](#Veri-Ön-İşleme)\n",
    "* [Topluluk Öğrenmesi](#Topluluk-Öğrenmesi-(Ensemble-Learning))\n",
    "    * [Basit Topluluk Teknikleri](#Basit-Topluluk-Teknikleri)\n",
    "        * [Max Voting](#Max-Voting)\n",
    "        * [Averaging](#Averaging)\n",
    "        * [Weighted Average](#Weighted-Average)\n",
    "    * [Gelişmiş Topluluk Teknikleri](#Gelişmiş-Topluluk-Teknikleri)\n",
    "        * [Stacking](#Stacking)\n",
    "        * [Blending](#Blending)\n",
    "        * [Bagging](#Bagging)\n",
    "        * [Boosting](#Boosting)\n",
    "    * [Bagging ve Boosting Kullanan Algoritmalar](#Bagging-ve-Boosting-Kullanan-Algoritmalar)\n",
    "        * [Bagging meta-estimator](#Bagging-meta-estimator)\n",
    "        * [Random Forest](#Random-Forest)\n",
    "        * [Gradient Boosting](#Gradient-Boosting)\n",
    "        * [XGBoost](#XGBoost)\n",
    "        * [LightGBM](#LightGBM)\n",
    "        * [CatBoost](#CatBoost)\n",
    "* [Sonuç](#Sonuç:)\n",
    "* [Kaynaklar](#Kaynaklar)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Veri Seti"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Veri Seti Tanımı\n",
    "\n",
    "15 Nisan 1912'de, ilk yolculuğu sırasında Titanic bir buzdağına çarparak battı ve 2224 yolcu ve mürettebattan 1502'sini öldürdü. Bu sansasyonel trajedi, uluslararası toplumu şok etti ve gemiler için daha iyi güvenlik düzenlemelerine yol açtı. Kazanın bu kadar can kaybına yol açmasının nedenlerinden biri, yolcular ve mürettebat için yeterli cankurtaran botunun olmamasıydı. Batarken hayatta kalmanın bir takım şans unsuru olmasına rağmen, kadınlar, çocuklar ve üst sınıflar gibi bazı grupların hayatta kalma olasılığı diğerlerinden daha fazlaydı. Bu olayda, ne tür insanların hayatta kalma olasılığının yüksek olduğunu tahmin edeceğiz. Özellikle, trajediden hangi yolcuların hayatta kaldığını tahmin etmek için makine öğrenimi araçlarını uygulayacağız."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hedef"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Yolcuların kaza sonucu yaşayıp yaşamadığını tahmin edelim. Hedef değişkenimiz 'Survived'. Hayatta kalanlar 1 ölenler 0 değerini alıyor.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Değişkenlerin Tanımları"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Survived : Hayatta kalan (1) ya da ölen(0)\n",
    "* Pclass : Yolcu sınıfı\n",
    "* Name : Yolcu ismi\n",
    "* Sex : Yolcu cinsiyeti\n",
    "* Age : Yolcu yaşı\n",
    "* SibSp : Gemideki kardeş/eş sayısı\n",
    "* Parch : Gemideki ebeveyn/çocuk sayısı\n",
    "- Ticket : Ticket Number\n",
    "- Fare : Ücret\n",
    "- Cabin : Kabin\n",
    "- Embarked : Liman"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gerekli Kütüphanelerin Kurulumu ve Veri Setinin Yüklenmesi"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>PassengerId</th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Name</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Ticket</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Cabin</th>\n",
       "      <th>Embarked</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Braund, Mr. Owen Harris</td>\n",
       "      <td>male</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>A/5 21171</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Cumings, Mrs. John Bradley (Florence Briggs Th...</td>\n",
       "      <td>female</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>PC 17599</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>C85</td>\n",
       "      <td>C</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>Heikkinen, Miss. Laina</td>\n",
       "      <td>female</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>STON/O2. 3101282</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>Futrelle, Mrs. Jacques Heath (Lily May Peel)</td>\n",
       "      <td>female</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>113803</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>C123</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>Allen, Mr. William Henry</td>\n",
       "      <td>male</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>373450</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>NaN</td>\n",
       "      <td>S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   PassengerId  Survived  Pclass  \\\n",
       "0            1         0       3   \n",
       "1            2         1       1   \n",
       "2            3         1       3   \n",
       "3            4         1       1   \n",
       "4            5         0       3   \n",
       "\n",
       "                                                Name     Sex   Age  SibSp  \\\n",
       "0                            Braund, Mr. Owen Harris    male  22.0      1   \n",
       "1  Cumings, Mrs. John Bradley (Florence Briggs Th...  female  38.0      1   \n",
       "2                             Heikkinen, Miss. Laina  female  26.0      0   \n",
       "3       Futrelle, Mrs. Jacques Heath (Lily May Peel)  female  35.0      1   \n",
       "4                           Allen, Mr. William Henry    male  35.0      0   \n",
       "\n",
       "   Parch            Ticket     Fare Cabin Embarked  \n",
       "0      0         A/5 21171   7.2500   NaN        S  \n",
       "1      0          PC 17599  71.2833   C85        C  \n",
       "2      0  STON/O2. 3101282   7.9250   NaN        S  \n",
       "3      0            113803  53.1000  C123        S  \n",
       "4      0            373450   8.0500   NaN        S  "
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# !pip install scikit-learn==0.21rc2\n",
    "import pandas as pd # veri analizi araçları\n",
    "import numpy as np # hesaplama araçları\n",
    "from matplotlib import pyplot as plt #görselleştirme\n",
    "import seaborn as sns # görselleştirme\n",
    "import sklearn\n",
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn.neighbors import KNeighborsClassifier\n",
    "from sklearn.tree import DecisionTreeClassifier\n",
    "\n",
    "from xgboost import XGBClassifier\n",
    "\n",
    "import warnings\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "\n",
    "df = pd.read_csv('titanic.csv',sep=',')\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Veri Setine Genel Bakış"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(891, 12)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape  # df setinin boyutları"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['PassengerId', 'Survived', 'Pclass', 'Name', 'Sex', 'Age', 'SibSp',\n",
       "       'Parch', 'Ticket', 'Fare', 'Cabin', 'Embarked'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.columns # değişkenler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>PassengerId</th>\n",
       "      <td>891.0</td>\n",
       "      <td>446.000000</td>\n",
       "      <td>257.353842</td>\n",
       "      <td>1.00</td>\n",
       "      <td>223.5000</td>\n",
       "      <td>446.0000</td>\n",
       "      <td>668.5</td>\n",
       "      <td>891.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Survived</th>\n",
       "      <td>891.0</td>\n",
       "      <td>0.383838</td>\n",
       "      <td>0.486592</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Pclass</th>\n",
       "      <td>891.0</td>\n",
       "      <td>2.308642</td>\n",
       "      <td>0.836071</td>\n",
       "      <td>1.00</td>\n",
       "      <td>2.0000</td>\n",
       "      <td>3.0000</td>\n",
       "      <td>3.0</td>\n",
       "      <td>3.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Age</th>\n",
       "      <td>714.0</td>\n",
       "      <td>29.699118</td>\n",
       "      <td>14.526497</td>\n",
       "      <td>0.42</td>\n",
       "      <td>20.1250</td>\n",
       "      <td>28.0000</td>\n",
       "      <td>38.0</td>\n",
       "      <td>80.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>SibSp</th>\n",
       "      <td>891.0</td>\n",
       "      <td>0.523008</td>\n",
       "      <td>1.102743</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>1.0</td>\n",
       "      <td>8.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Parch</th>\n",
       "      <td>891.0</td>\n",
       "      <td>0.381594</td>\n",
       "      <td>0.806057</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>6.0000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Fare</th>\n",
       "      <td>891.0</td>\n",
       "      <td>32.204208</td>\n",
       "      <td>49.693429</td>\n",
       "      <td>0.00</td>\n",
       "      <td>7.9104</td>\n",
       "      <td>14.4542</td>\n",
       "      <td>31.0</td>\n",
       "      <td>512.3292</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             count        mean         std   min       25%       50%    75%  \\\n",
       "PassengerId  891.0  446.000000  257.353842  1.00  223.5000  446.0000  668.5   \n",
       "Survived     891.0    0.383838    0.486592  0.00    0.0000    0.0000    1.0   \n",
       "Pclass       891.0    2.308642    0.836071  1.00    2.0000    3.0000    3.0   \n",
       "Age          714.0   29.699118   14.526497  0.42   20.1250   28.0000   38.0   \n",
       "SibSp        891.0    0.523008    1.102743  0.00    0.0000    0.0000    1.0   \n",
       "Parch        891.0    0.381594    0.806057  0.00    0.0000    0.0000    0.0   \n",
       "Fare         891.0   32.204208   49.693429  0.00    7.9104   14.4542   31.0   \n",
       "\n",
       "                  max  \n",
       "PassengerId  891.0000  \n",
       "Survived       1.0000  \n",
       "Pclass         3.0000  \n",
       "Age           80.0000  \n",
       "SibSp          8.0000  \n",
       "Parch          6.0000  \n",
       "Fare         512.3292  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe().T #betimsel istatistikler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PassengerId      int64\n",
       "Survived         int64\n",
       "Pclass           int64\n",
       "Name            object\n",
       "Sex             object\n",
       "Age            float64\n",
       "SibSp            int64\n",
       "Parch            int64\n",
       "Ticket          object\n",
       "Fare           float64\n",
       "Cabin           object\n",
       "Embarked        object\n",
       "dtype: object"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.dtypes #değişkenlerin veri tipler"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Bu değişkenlerden Cabin ve Ticket string tipinde ifade edilmiştir.\n",
    "\n",
    "Sex değişkeni binary category tipindedir Embarked değişkeni category tipindedir Name değişkeni text tipindedir\n",
    "\n",
    "Pclass, Age, PassengerID, SibSp, Parch, Fare olmak üzere 6 değişken nümeric türündendir<br>\n",
    "\n",
    "Age ve Fare değişkenleri sürekli sayısal Pclass değişkeni ordinal category tipindedir \n",
    "\n",
    "SibSp ve Parch değişkenleri discrete ordinal Hedef değişken olan Survived değişkeni binary categorical"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Hayatta kalma oranı"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.3838383838383838"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Survived'].mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Cinsiyete Göre Hayatta Kalma Oranı"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Sex</th>\n",
       "      <th>Survived</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>female</td>\n",
       "      <td>0.742038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>male</td>\n",
       "      <td>0.188908</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Sex  Survived\n",
       "0  female  0.742038\n",
       "1    male  0.188908"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[[\"Sex\", \"Survived\"]].groupby(['Sex'], as_index=False).mean().sort_values(by='Survived', ascending=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Veri Setinin Görselleştirilmesi\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8,4))\n",
    "Survived = df[\"Survived\"].value_counts()\n",
    "sns.barplot(y=Survived.values, x=Survived.index, alpha=0.6)\n",
    "plt.title(\"'Survived' Değişkeninin Frekans Değerleri\")\n",
    "plt.xlabel('Yolcuların Hayatta Kalma Durumu', fontsize=10)\n",
    "plt.ylabel('Frekans', fontsize=10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(df['Embarked'], df['Survived'])\n",
    "plt.title(\"Yolcuların Liman Türüne Göre Hayatta Kalma Durumu\")\n",
    "plt.xlabel('Liman Tipi', fontsize=10)\n",
    "plt.ylabel('Hayatta Kalma Durumu', fontsize=10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(df['Pclass'], df['Survived'])\n",
    "plt.title(\"Yolcu Sınıfına Göre Hayatta Kalma Durumu\")\n",
    "plt.xlabel('Yolcu Sınıfı', fontsize=10)\n",
    "plt.ylabel('Hayatta Kalma Durumu', fontsize=10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.barplot(df['Sex'], df['Survived'])\n",
    "plt.title(\"Cinsiyete Göre Hayatta Kalma Durumu\")\n",
    "plt.xlabel('Cinsiyet', fontsize=10)\n",
    "plt.ylabel('Hayatta Kalma Durumu', fontsize=10)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Kayıp Değer Analizi\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['Age', 'Cabin', 'Embarked'], dtype='object')"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# veri setinde eksik değerleri olan sütunları listele\n",
    "df.columns[df.isna().any()]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#pip install missingno\n",
    "import missingno as msno\n",
    "msno.matrix(df);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "PassengerId      0\n",
       "Survived         0\n",
       "Pclass           0\n",
       "Name             0\n",
       "Sex              0\n",
       "Age            177\n",
       "SibSp            0\n",
       "Parch            0\n",
       "Ticket           0\n",
       "Fare             0\n",
       "Cabin          687\n",
       "Embarked         2\n",
       "dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isna().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# eksik değerleri Embarked için mode ile doldurma\n",
    "embarked = df[\"Embarked\"].mode()[0]\n",
    "df[\"Embarked\"] = df[\"Embarked\"].fillna(embarked)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cabin değişkeninin çok büyük oranda eksik değer içermesinden dolayı veri setinden silindi.\n",
    "\n",
    "df.drop(\"Cabin\", axis = 1, inplace = True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Age değişkeni için çeşitli veri doldurma yöntemlerini uygulayalım.\n",
    "# ilk olarak dağılımına bakalım\n",
    "sns.distplot(df[\"Age\"])\n",
    "plt.title(\"'Age' Değişkeninin Dağılımı\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# mean ile dolduralım\n",
    "df_mean = df.copy()\n",
    "mean_age = df[\"Age\"].mean() # ortalama yaş değeri\n",
    "df_mean['Age'] = df_mean[\"Age\"].fillna(np.round(mean_age,1))\n",
    "\n",
    "sns.distplot(df[\"Age\"],color=\"olive\", label=\"Orijinal Hali\")\n",
    "sns.distplot( df_mean[\"Age\"],color=\"teal\", label=\"Ort. İle Düzenlenmiş Hali\")\n",
    "plt.title(\"Orijinal 'Age' vs Ort. İle Düzenlenmiş 'Age' \")\n",
    "plt.legend()\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# median ile dolduralım\n",
    "\n",
    "df_median = df.copy()\n",
    "median_age = df[\"Age\"].median()\n",
    "df_median[\"Age\"] = df_median[\"Age\"].fillna(median_age)\n",
    "\n",
    "sns.distplot(df[\"Age\"] , color=\"red\", label=\"Orijinal Hali\")\n",
    "sns.distplot(df_median[\"Age\"] , color=\"skyblue\", label=\"Medyan İle Düzenlenmiş Hali\")\n",
    "plt.title(\"Orijinal 'Age' vs Medyan İle Düzenlenmiş 'Age' \")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# mod ile dolduralım\n",
    "df_mode = df.copy()\n",
    "age_mode = df[\"Age\"].mode()[0] # modunu bulma\n",
    "df_mode[\"Age\"] = df_mode[\"Age\"].fillna(age_mode)\n",
    "\n",
    "sns.distplot(df[\"Age\"] , color=\"red\", label=\"Orijinal Hali\")\n",
    "sns.distplot(df_mode[\"Age\"] , color=\"skyblue\", label=\"Mod İle Düzenlenmiş Hali\")\n",
    "plt.title(\"Orijinal 'Age' vs Mod İle Düzenlenmiş 'Age' \")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Veri dağılımından harektle mean ile doldurmaya karar verildi.\n",
    "\n",
    "df[\"Age\"] = df[\"Age\"].fillna(np.round(mean_age,1))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Veri Ön İşleme\n",
    "\n",
    "### Kategorik değişkeni sayısal değerlere dönüştürme"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 1 CİNSİYET DEĞİŞKENİ İÇİN ANALİZ\n",
    "# Binary encoding\n",
    "\n",
    "df[\"Sex\"] = df[\"Sex\"].map({\"male\":0, \"female\":1}).astype(int)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Southampton(S)'dan binen yolcu sayısı: 646\n",
      "Cherbourg (C)'dan binen yolcu sayısı: 168\n",
      "Queenstown(Q)'dan binen yolcu sayısı: 77\n"
     ]
    }
   ],
   "source": [
    "# EMBARKED Değişkeni İçin Analiz\n",
    "\n",
    "southampton = df[df[\"Embarked\"] == \"S\"].shape[0]\n",
    "print(\"Southampton(S)'dan binen yolcu sayısı:\",southampton)\n",
    "\n",
    "cherbourg = df[df[\"Embarked\"] == \"C\"].shape[0]\n",
    "print(\"Cherbourg (C)'dan binen yolcu sayısı:\",cherbourg)\n",
    "\n",
    "queenstown = df[df[\"Embarked\"] == \"Q\"].shape[0]\n",
    "print(\"Queenstown(Q)'dan binen yolcu sayısı:\",queenstown )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import OneHotEncoder\n",
    "\n",
    "# embarked değişkeni ordinal bir yapıda olmadığından one hot encoder kullanılabilir.\n",
    "\n",
    "encoder = OneHotEncoder()\n",
    "encoder_df = pd.DataFrame(encoder.fit_transform(df[[\"Embarked\"]]).toarray()).add_prefix('Embarked_')\n",
    "df = df.join(encoder_df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array(['C', 'Q', 'S'], dtype=object)]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "encoder.categories_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Name Değişkeni İçin Analiz:\n",
    "import re"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_title(df):\n",
    "    for dataset in df:\n",
    "        dataset['Title'] = dataset.Name.str.extract('([A-Za-z]+)\\.', expand=False)\n",
    "\n",
    "make_title([df])    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Mr', 'Mrs', 'Miss', 'Master', 'Don', 'Rev', 'Dr', 'Mme', 'Ms',\n",
       "       'Major', 'Lady', 'Sir', 'Mlle', 'Col', 'Capt', 'Countess',\n",
       "       'Jonkheer'], dtype=object)"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.Title.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>Sex</th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Title</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>Capt</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Col</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Countess</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Don</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dr</th>\n",
       "      <td>6</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Jonkheer</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lady</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Major</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Master</th>\n",
       "      <td>40</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Miss</th>\n",
       "      <td>0</td>\n",
       "      <td>182</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mlle</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mme</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mr</th>\n",
       "      <td>517</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Mrs</th>\n",
       "      <td>0</td>\n",
       "      <td>125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Ms</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Rev</th>\n",
       "      <td>6</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Sir</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Sex         0    1\n",
       "Title             \n",
       "Capt        1    0\n",
       "Col         2    0\n",
       "Countess    0    1\n",
       "Don         1    0\n",
       "Dr          6    1\n",
       "Jonkheer    1    0\n",
       "Lady        0    1\n",
       "Major       2    0\n",
       "Master     40    0\n",
       "Miss        0  182\n",
       "Mlle        0    2\n",
       "Mme         0    1\n",
       "Mr        517    0\n",
       "Mrs         0  125\n",
       "Ms          0    1\n",
       "Rev         6    0\n",
       "Sir         1    0"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.crosstab(df['Title'], df['Sex'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Veri setinde yer alan name değişkeninde bulanan unvanları daha genel kullanımına \n",
    "# göre replace metoduyla değiştirelim ve hayatta kalma durumuna göre gruplayalım\n",
    "\n",
    "def Title_arange(df):\n",
    "    for dataset in df:\n",
    "        dataset['Title'] = dataset['Title'].replace(['Lady', 'Countess','Capt', 'Col',\\\n",
    "     'Don', 'Dr', 'Major', 'Rev', 'Sir', 'Jonkheer', 'Dona'], 'Rare')\n",
    "\n",
    "        dataset['Title'] = dataset['Title'].replace('Mlle', 'Miss')\n",
    "        dataset['Title'] = dataset['Title'].replace('Ms', 'Miss')\n",
    "        dataset['Title'] = dataset['Title'].replace('Mme', 'Mrs')\n",
    "\n",
    "        #train_df[['Title', 'Survived']].groupby(['Title'], as_index=False).mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "Title_arange([df])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "encoder = OneHotEncoder()\n",
    "\n",
    "encoder_df = pd.DataFrame(encoder.fit_transform(df[[\"Title\"]]).toarray()).add_prefix('Title_')\n",
    "df = df.join(encoder_df)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Daha sonra Name, Embarked, Ticket değişkenini silelim\n",
    "\n",
    "\n",
    "df = df.drop([\"PassengerId\",\"Name\",\"Ticket\",\"Embarked\",\"Title\"], axis= 1)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array(['Master', 'Miss', 'Mr', 'Mrs', 'Rare'], dtype=object)]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "encoder.categories_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Survived</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Embarked_0</th>\n",
       "      <th>Embarked_1</th>\n",
       "      <th>Embarked_2</th>\n",
       "      <th>Title_0</th>\n",
       "      <th>Title_1</th>\n",
       "      <th>Title_2</th>\n",
       "      <th>Title_3</th>\n",
       "      <th>Title_4</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>22.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7.2500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>38.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>71.2833</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>26.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>53.1000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>35.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>8.0500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Survived  Pclass  Sex   Age  SibSp  Parch     Fare  Embarked_0  Embarked_1  \\\n",
       "0         0       3    0  22.0      1      0   7.2500         0.0         0.0   \n",
       "1         1       1    1  38.0      1      0  71.2833         1.0         0.0   \n",
       "2         1       3    1  26.0      0      0   7.9250         0.0         0.0   \n",
       "3         1       1    1  35.0      1      0  53.1000         0.0         0.0   \n",
       "4         0       3    0  35.0      0      0   8.0500         0.0         0.0   \n",
       "\n",
       "   Embarked_2  Title_0  Title_1  Title_2  Title_3  Title_4  \n",
       "0         1.0      0.0      0.0      1.0      0.0      0.0  \n",
       "1         0.0      0.0      0.0      0.0      1.0      0.0  \n",
       "2         1.0      0.0      1.0      0.0      0.0      0.0  \n",
       "3         1.0      0.0      0.0      0.0      1.0      0.0  \n",
       "4         1.0      0.0      0.0      1.0      0.0      0.0  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### VERİ SETİNİ TRAİN TEST VE VALİDATİON OLARAK AYIRALIM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "y = df[\"Survived\"]\n",
    "\n",
    "X = df.drop([\"Survived\"], axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=1)\n",
    "\n",
    "X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size=0.25, random_state=44) \n",
    "X_train.reset_index(inplace=True,drop=True)\n",
    "X_test.reset_index(inplace=True,drop=True)\n",
    "X_val.reset_index(inplace=True,drop=True)\n",
    "y_train.reset_index(inplace=True,drop=True)\n",
    "y_test.reset_index(inplace=True,drop=True)\n",
    "y_val.reset_index(inplace=True,drop=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Topluluk Öğrenmesi (Ensemble Learning)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Topluluk öğrenme kavramını bir örnekle anlayalım. Bir film yönetmeni olduğunuzu ve çok önemli ve ilginç bir konu hakkında kısa bir film hazırladığınızı varsayalım. Şimdi, filmi herkese açık hale getirmeden önce ön geri bildirim (derecelendirme) almak istiyorsunuz. Bunu yapmanın olası yolları nelerdir?\n",
    "\n",
    "* A: Bir arkadaşınızdan filmi sizin için değerlendirmesini isteyebilirsiniz.\n",
    "\n",
    "Bu durumda arkadaşınızın sizi çok sevmesi ve yarattığınız korkunç çalışmaya 1 yıldız vererek kalbinizi kırmak istememesi tamamen mümkün.\n",
    "\n",
    "* B: Başka bir yol da 5 meslektaşınızdan filmi derecelendirmelerini istemek olabilir.\n",
    "\n",
    "Bu yöntem, filminiz için dürüst derecelendirmeler sağlayabilir. Ama hala bir sorun var. Bu 5 kişi, filminizin konusu için “Konu Uzmanı” olmayabilir.\n",
    "\n",
    "* C: Ya 50 kişiden filmi değerlendirmesini istersek ?\n",
    "\n",
    "Bazıları arkadaşınız olabilir, bazıları meslektaşlarınız olabilir ve hatta bazıları tamamen yabancı olabilir. Bu durumda yanıtlar, artık farklı becerilere sahip insanlara sahip olduğunuz için daha genel ve çeşitli olacaktır. Bu örneklerle, farklı bir grup insanın, bireylere kıyasla daha iyi kararlar verebileceği sonucuna varabilirsiniz. Tekli modellere kıyasla çok çeşitli modeller için benzer durum geçerlidir. Makine Öğrenimindeki bu çeşitlilik, Topluluk Öğrenimi adı verilen bir teknikle elde edilir. Artık topluluk öğreniminin ne olduğu konusunda bir fikir edindiğinize göre, bunların uygulamalarıyla birlikte topluluk öğrenmedeki çeşitli tekniklere bakalım."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](elephant.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Basit Topluluk Teknikleri"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Bu bölümde, birkaç basit ama güçlü tekniğe bakacağız, yani:\n",
    "    \n",
    "* Max Voting\n",
    "* Averaging\n",
    "* Weighted Averaging"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Max Voting\n",
    "\n",
    "Maksimum oylama yöntemi genellikle sınıflandırma problemleri için kullanılır. Bu teknikte, her veri noktası için tahmin yapmak için birden çok model kullanılır. Her modelin tahminleri bir 'oy' olarak kabul edilir. Modellerin çoğundan aldığımız tahminler son tahmin olarak kullanılmaktadır.\n",
    "\n",
    "Örneğin, 5 meslektaşınızdan filminizi derecelendirmesini istediğinizde (5 üzerinden); Bunlardan üçünün 4, ikisinin de 5 verdiğini varsayacağız. Çoğunluk 4 puan verdiği için final derecesini 4 olarak alabiliriz. Bunu tüm tahminlerin modunu almak olarak düşünebilirsiniz."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](maxvoting.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Örnek Kod:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1, 0, 1, 1, 1, 0, 0, 1, 1, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0,\n",
       "       1, 0, 0, 0, 1, 1, 0, 1, 0, 1, 0, 0, 0, 0, 1, 0, 1, 0, 1, 1, 1, 1,\n",
       "       1, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0,\n",
       "       0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0,\n",
       "       0, 1, 0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0,\n",
       "       0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 1, 1, 0, 0, 0, 0, 0,\n",
       "       1, 1, 0, 1, 1, 0, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0,\n",
       "       0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 1, 0,\n",
       "       0, 0, 1], dtype=int64)"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import VotingClassifier\n",
    "\n",
    "model1 = LogisticRegression(random_state=10)\n",
    "model2 = DecisionTreeClassifier(random_state=10)\n",
    "model3 = KNeighborsClassifier()\n",
    "\n",
    "model = VotingClassifier(estimators=[('lr', model1), ('dt', model2),('kn', model3)], voting='hard')\n",
    "model.fit(X_train,y_train)\n",
    "model.predict(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Averaging"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Maksimum oylama tekniğine benzer şekilde, ortalamada her veri noktası için birden fazla tahmin yapılır. Bu yöntemde, tüm modellerden ortalama bir tahmin alıyoruz ve bunu son tahmini yapmak için kullanıyoruz. Ortalama alma, regresyon problemlerinde tahmin yapmak için veya sınıflandırma problemleri için olasılıkları hesaplarken kullanılabilir.\n",
    "\n",
    "Örneğin, aşağıdaki durumda, ortalama alma yöntemi tüm değerlerin ortalamasını alacaktır. \n",
    "\n",
    "yani (5 + 4 + 5 + 4 + 4) / 5 = 4.4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](averaging.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Örnek Kod:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[1.         0.         0.66666667 1.         1.         0.\n",
      " 0.         1.         0.66666667 0.66666667 0.         0.33333333\n",
      " 0.         0.         1.         0.33333333 0.         0.\n",
      " 0.         1.         0.         0.         0.66666667 0.\n",
      " 0.33333333 0.33333333 1.         1.         0.         1.\n",
      " 0.33333333 0.66666667 0.33333333 0.         0.         0.\n",
      " 0.66666667 0.         0.66666667 0.         0.66666667 0.66666667\n",
      " 0.66666667 0.66666667 1.         0.         0.         0.\n",
      " 1.         0.         0.         0.66666667 0.33333333 0.\n",
      " 0.         0.66666667 0.         0.         0.         0.33333333\n",
      " 0.         0.         0.66666667 0.66666667 0.         0.33333333\n",
      " 0.         0.         1.         0.         0.         0.\n",
      " 0.66666667 0.         0.66666667 0.33333333 1.         0.\n",
      " 0.         1.         0.33333333 0.         0.         0.\n",
      " 0.66666667 0.         0.         0.         0.         1.\n",
      " 0.33333333 0.         0.         1.         0.         0.66666667\n",
      " 0.         0.66666667 0.         0.         1.         0.33333333\n",
      " 0.66666667 0.33333333 1.         0.         0.         0.\n",
      " 0.         0.         0.33333333 0.33333333 0.         0.\n",
      " 0.         0.         0.33333333 0.33333333 1.         0.33333333\n",
      " 1.         0.33333333 0.         0.         1.         1.\n",
      " 0.66666667 0.33333333 0.         0.         0.         0.\n",
      " 1.         1.         0.         1.         0.66666667 0.\n",
      " 0.         0.66666667 1.         0.         1.         0.66666667\n",
      " 0.         0.33333333 0.         0.         0.33333333 0.\n",
      " 1.         0.         0.         0.33333333 0.33333333 0.\n",
      " 0.         0.         0.66666667 0.         0.         0.\n",
      " 0.33333333 0.66666667 0.         0.33333333 1.         0.33333333\n",
      " 0.33333333 0.33333333 1.         0.         1.         0.33333333\n",
      " 0.66666667 0.         0.         0.         1.        ]\n"
     ]
    }
   ],
   "source": [
    "model1 = DecisionTreeClassifier()\n",
    "model2 = KNeighborsClassifier()\n",
    "model3= LogisticRegression()\n",
    "\n",
    "model1.fit(X_train,y_train)\n",
    "model2.fit(X_train,y_train)\n",
    "model3.fit(X_train,y_train)\n",
    "\n",
    "pred1=model1.predict(X_test)\n",
    "pred2=model2.predict(X_test)\n",
    "pred3=model3.predict(X_test)\n",
    "\n",
    "finalpred=(pred1+pred2+pred3)/3\n",
    "print(finalpred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Weighted Average\n",
    "\n",
    "Bu, ortalama alma yönteminin bir uzantısıdır. Tüm modellere, tahmin için her modelin önemini tanımlayan farklı ağırlıklar atanmıştır. Örneğin, meslektaşlarınızdan ikisi eleştirmen ise, diğerlerinin bu alanda daha önceden deneyimi yoksa, bu iki arkadaşın cevaplarına diğer insanlara göre daha fazla önem verilir.\n",
    "\n",
    "Sonuç [(5 * 0.23) + (4 * 0.23) + (5 * 0.18) + (4 * 0.18) + (4 * 0.18)] = 4.41 olarak hesaplanır"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](weighted.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### Örnek Kod:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([1. , 0. , 0.7, 1. , 1. , 0. , 0. , 1. , 0.6, 0.7, 0. , 0.4, 0. ,\n",
       "       0. , 1. , 0.3, 0. , 0. , 0. , 1. , 0. , 0. , 0.7, 0. , 0.4, 0.3,\n",
       "       1. , 1. , 0. , 1. , 0.4, 0.6, 0.3, 0. , 0. , 0. , 0.6, 0. , 0.6,\n",
       "       0. , 0.7, 0.7, 0.7, 0.6, 1. , 0. , 0. , 0. , 1. , 0.3, 0. , 0.7,\n",
       "       0.3, 0. , 0. , 0.7, 0. , 0. , 0. , 0.4, 0. , 0. , 0.3, 0.6, 0. ,\n",
       "       0.4, 0. , 0. , 1. , 0. , 0. , 0. , 0.6, 0. , 0.7, 0.3, 1. , 0. ,\n",
       "       0. , 1. , 0.3, 0. , 0. , 0. , 0.7, 0. , 0. , 0. , 0. , 1. , 0.3,\n",
       "       0. , 0. , 1. , 0. , 0.7, 0. , 0.7, 0. , 0. , 1. , 0.4, 0.7, 0.4,\n",
       "       1. , 0. , 0. , 0. , 0. , 0. , 0.3, 0.3, 0. , 0. , 0. , 0. , 0.3,\n",
       "       0.3, 1. , 0.4, 1. , 0.3, 0. , 0. , 1. , 1. , 0.7, 0.4, 0. , 0. ,\n",
       "       0. , 0. , 1. , 1. , 0. , 1. , 0.7, 0.3, 0. , 0.7, 1. , 0. , 1. ,\n",
       "       0.7, 0. , 0.4, 0. , 0. , 0.3, 0. , 1. , 0. , 0. , 0.4, 0.3, 0. ,\n",
       "       0. , 0. , 0.7, 0. , 0. , 0. , 0.4, 0.6, 0. , 0.4, 1. , 0.3, 0.3,\n",
       "       0.3, 1. , 0.3, 1. , 0.3, 0.7, 0. , 0. , 0. , 1. ])"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model1 = DecisionTreeClassifier()\n",
    "model2 = KNeighborsClassifier()\n",
    "model3= LogisticRegression()\n",
    "\n",
    "model1.fit(X_train,y_train)\n",
    "model2.fit(X_train,y_train)\n",
    "model3.fit(X_train,y_train)\n",
    "\n",
    "pred1=model1.predict(X_test)\n",
    "pred2=model2.predict(X_test)\n",
    "pred3=model3.predict(X_test)\n",
    "\n",
    "finalpred=(pred1*0.3+pred2*0.3+pred3*0.4)\n",
    "finalpred"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Gelişmiş Topluluk Teknikleri\n",
    "\n",
    "Artık temel topluluk tekniklerini ele aldığımıza göre, gelişmiş teknikleri anlamaya geçelim."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Stacking\n",
    "\n",
    "Yığınlama, yeni bir model oluşturmak için birden çok modelden (örneğin karar ağacı, knn veya svm) gelen tahminleri kullanan bir toplu öğrenme tekniğidir. Bu model, test seti üzerinde tahmin yapmak için kullanılır. Aşağıda basit bir yığılmış topluluk için adım adım bir açıklama bulunmaktadır:\n",
    "\n",
    "1. Eğitim verisi 10 parçaya bölünmüştür\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](stacking1.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. Bir temel model (örneğin karar ağacı) 9 parçaya oturtulur ve 10. parça için tahminler yapılır. Bu, eğitim setinin her bölümü için yapılır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](stacking2.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3. Temel model (bu durumda karar ağacı) daha sonra tüm eğitim veri setine yerleştirilir.\n",
    "\n",
    "4. Bu model kullanılarak test seti üzerinde tahminler yapılır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](stacking3.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "5. 2'den 4'e kadar olan adımlar başka bir temel model (knn gibi) için tekrar edilir ve bu da eğitim seti ve test seti için başka bir tahmin seti ile sonuçlanır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](stacking4.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "6. Eğitim setindeki tahminler, oluşacak yeni bir modeli etkiler."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](stacking5.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "7. Bu model, test tahmin seti üzerinde son tahminlerde bulunmak için kullanılır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Örnek Kod:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "def Stacking(model,train,y,test,n_fold):\n",
    "    \n",
    "    folds=StratifiedKFold(n_splits=n_fold,random_state=1)\n",
    "    test_pred=np.empty((test.shape[0],1),float)\n",
    "    train_pred=np.empty((0,1),float)\n",
    "    for train_indices,val_indices in folds.split(train,y.values):\n",
    "        x_train,x_val=train.iloc[train_indices],train.iloc[val_indices]\n",
    "        y_train,y_val=y.iloc[train_indices],y.iloc[val_indices]\n",
    "\n",
    "        model.fit(X=x_train,y=y_train)\n",
    "        train_pred=np.append(train_pred,model.predict(x_val))\n",
    "        test_pred=np.append(test_pred,model.predict(test))\n",
    "    return test_pred.reshape(-1,1),train_pred\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import StratifiedKFold\n",
    "\n",
    "model1 = DecisionTreeClassifier(random_state=10)\n",
    "\n",
    "test_pred1 ,train_pred1=Stacking(model=model1,n_fold=10, train=X_train,test=X_test,y=y_train)\n",
    "\n",
    "train_pred1=pd.DataFrame(train_pred1)\n",
    "test_pred1=pd.DataFrame(test_pred1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>529</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>530</th>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>531</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>532</th>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>533</th>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>534 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       0\n",
       "0    0.0\n",
       "1    0.0\n",
       "2    0.0\n",
       "3    0.0\n",
       "4    0.0\n",
       "..   ...\n",
       "529  1.0\n",
       "530  0.0\n",
       "531  1.0\n",
       "532  0.0\n",
       "533  1.0\n",
       "\n",
       "[534 rows x 1 columns]"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_pred1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Blending\n",
    "\n",
    "Blending, stacking ile aynı yaklaşımı izler, ancak tahmin yapmak için eğitim setinden yalnızca bir doğrulama (validation) seti kullanır. Diğer bir deyişle, stackingden farklı olarak, tahminler yalnızca validation setinde yapılır. Validation seti ve tahminler, test setinde çalıştırılan bir model oluşturmak için kullanılır. İşte blending işleminin ayrıntılı bir açıklaması:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Eğitim veri seti, tekrardan eğitim ve doğrulama (validation) setlerine ayrılmıştır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](blending1.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2. Modeller eğitim setine uygulanır.\n",
    "3. Tahminler, doğrulama seti ve test seti üzerinde yapılır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](blending2.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "4. Doğrulama setindeki tahminler, oluşacak yeni bir modeli etkiler.\n",
    "5. Bu model, test ve meta özellikler hakkında son tahminlerde bulunmak için kullanılır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Örnek Kod:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "model1 = DecisionTreeClassifier()\n",
    "model1.fit(X_train, y_train)\n",
    "val_pred1=model1.predict(X_val)\n",
    "test_pred1=model1.predict(X_test)\n",
    "val_pred1=pd.DataFrame(val_pred1)\n",
    "test_pred1=pd.DataFrame(test_pred1)\n",
    "\n",
    "model2 = KNeighborsClassifier()\n",
    "model2.fit(X_train,y_train)\n",
    "val_pred2=model2.predict(X_val)\n",
    "test_pred2=model2.predict(X_test)\n",
    "val_pred2=pd.DataFrame(val_pred2)\n",
    "test_pred2=pd.DataFrame(test_pred2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_val.reset_index(inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_val=pd.concat([X_val, val_pred1,val_pred2],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>Pclass</th>\n",
       "      <th>Sex</th>\n",
       "      <th>Age</th>\n",
       "      <th>SibSp</th>\n",
       "      <th>Parch</th>\n",
       "      <th>Fare</th>\n",
       "      <th>Embarked_0</th>\n",
       "      <th>Embarked_1</th>\n",
       "      <th>Embarked_2</th>\n",
       "      <th>Title_0</th>\n",
       "      <th>Title_1</th>\n",
       "      <th>Title_2</th>\n",
       "      <th>Title_3</th>\n",
       "      <th>Title_4</th>\n",
       "      <th>0</th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>34.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>19.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>91.0792</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "      <td>29.7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.7375</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>52.0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>93.5000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>173</th>\n",
       "      <td>173</td>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>34.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>13.0000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>174</th>\n",
       "      <td>174</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>29.7</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>27.7208</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>175</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>29.7</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>7.7500</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>176</th>\n",
       "      <td>176</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>39.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.9250</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>177</th>\n",
       "      <td>177</td>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>24.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>7.7958</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>178 rows × 17 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     index  Pclass  Sex   Age  SibSp  Parch     Fare  Embarked_0  Embarked_1  \\\n",
       "0        0       2    1  24.0      0      0  13.0000         0.0         0.0   \n",
       "1        1       2    1  34.0      0      0  13.0000         0.0         0.0   \n",
       "2        2       1    1  19.0      1      0  91.0792         1.0         0.0   \n",
       "3        3       3    1  29.7      0      0   7.7375         0.0         1.0   \n",
       "4        4       1    1  52.0      1      1  93.5000         0.0         0.0   \n",
       "..     ...     ...  ...   ...    ...    ...      ...         ...         ...   \n",
       "173    173       2    0  34.0      0      0  13.0000         0.0         0.0   \n",
       "174    174       1    0  29.7      0      0  27.7208         1.0         0.0   \n",
       "175    175       3    0  29.7      1      0   7.7500         0.0         1.0   \n",
       "176    176       3    0  39.0      0      0   7.9250         0.0         0.0   \n",
       "177    177       3    0  24.0      0      0   7.7958         0.0         0.0   \n",
       "\n",
       "     Embarked_2  Title_0  Title_1  Title_2  Title_3  Title_4  0  0  \n",
       "0           1.0      0.0      1.0      0.0      0.0      0.0  1  0  \n",
       "1           1.0      0.0      1.0      0.0      0.0      0.0  1  1  \n",
       "2           0.0      0.0      0.0      0.0      1.0      0.0  1  1  \n",
       "3           0.0      0.0      1.0      0.0      0.0      0.0  1  1  \n",
       "4           1.0      0.0      0.0      0.0      1.0      0.0  1  0  \n",
       "..          ...      ...      ...      ...      ...      ... .. ..  \n",
       "173         1.0      0.0      0.0      1.0      0.0      0.0  0  0  \n",
       "174         0.0      0.0      0.0      1.0      0.0      0.0  1  0  \n",
       "175         0.0      0.0      0.0      1.0      0.0      0.0  0  0  \n",
       "176         1.0      0.0      0.0      1.0      0.0      0.0  0  0  \n",
       "177         1.0      0.0      0.0      1.0      0.0      0.0  0  0  \n",
       "\n",
       "[178 rows x 17 columns]"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_val"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0      0\n",
       "1      1\n",
       "2      1\n",
       "3      1\n",
       "4      0\n",
       "      ..\n",
       "173    0\n",
       "174    0\n",
       "175    0\n",
       "176    0\n",
       "177    0\n",
       "Name: val_pred2, Length: 178, dtype: int64"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_val.iloc[:,-1].rename('val_pred2',inplace=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bagging\n",
    "\n",
    "Baggingin arkasındaki fikir, genelleştirilmiş bir sonuç elde etmek için birden fazla modelin (örneğin tüm karar ağaçları) sonuçlarını birleştirmektir. İşte bir soru: Tüm modelleri aynı veri kümesi üzerinde oluşturur ve birleştirirseniz, faydalı olur mu? Aynı girdiyi aldıkları için bu modellerin aynı sonucu verme şansı yüksektir. Peki bu sorunu nasıl çözebiliriz? Bunun için bootstrapping tekniğini ele alalım.\n",
    "\n",
    "Bootstrapping, orijinal veri kümesinden değiştirerek gözlemlerin alt kümelerini oluşturduğumuz bir örnekleme tekniğidir. Alt kümelerin boyutu, orijinal kümenin boyutuyla aynıdır.\n",
    "\n",
    "Bagging (veya Bootstrap Aggregating) tekniği, dağıtım (tam veri seti) hakkında adil bir fikir edinmek için bu alt kümeleri (bags) kullanır. Bagging için oluşturulan alt kümelerin boyutu, orijinal setten daha az olabilir."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](bagging1.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Orijinal veri kümesinden, yerine koyma uygulanması ile gözlemler seçilerek birden çok alt küme oluşturulur.\n",
    "\n",
    "2. Bu alt kümelerin her birinde bir temel model (zayıf model) oluşturulur.\n",
    "\n",
    "3. Modeller paralel çalışır ve birbirinden bağımsızdır.\n",
    "\n",
    "4. Nihai tahminler, tüm modellerden alınan tahminler birleştirilerek belirlenir."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](bagging2.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Boosting\n",
    "\n",
    "Daha ileri gitmeden önce, işte size başka bir soru: Bir veri noktası ilk model tarafından yanlış tahmin edilirse, bir sonraki adımda tüm modellerin bir arada çalışması tahminleri birleştirmek daha iyi sonuçlar sağlar mı? Bu tür durumlar için boosting kullanılır.\n",
    "\n",
    "Boosting, sonraki her modelin önceki modelin hatalarını düzeltmeye çalıştığı sıralı bir süreçtir. Sonraki modeller önceki modele bağlıdır. Aşağıdaki adımlarla boostingin nasıl çalıştığını anlayalım.\n",
    "\n",
    "1. Orijinal veri kümesinden bir alt küme oluşturulur.\n",
    "2. Başlangıçta, tüm veri noktalarına eşit ağırlıklar verilir.\n",
    "3. Bu alt kümede bir temel model oluşturulur.\n",
    "4. Bu model, tüm veri seti üzerinde tahminler yapmak için kullanılır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](boosting1.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "5. Hatalar, gerçek değerler ve tahmin edilen değerler kullanılarak hesaplanır.\n",
    "6. Yanlış tahmin edilen gözlemlere daha yüksek ağırlık verilir. (Burada, yanlış sınıflandırılmış üç mavi artı noktaya daha yüksek ağırlık verilecektir)\n",
    "7. Başka bir model oluşturulur ve veri kümesi üzerinde tahminler yapılır\n",
    "(Bu model, önceki modeldeki hataları düzeltmeye çalışır)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](boosting2.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "8. Benzer şekilde, her biri önceki modelin hatalarını düzelten birden çok model oluşturulur.\n",
    "9. Nihai model (güçlü öğrenci), tüm modellerin (zayıf öğrenci) ağırlıklı ortalamasıdır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](boosting3.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Dolayısıyla, boosting algoritması, güçlü bir öğrenci oluşturmak için bir dizi zayıf öğrenciyi birleştirir. Bireysel modeller, tüm veri kümesinde iyi performans göstermez, ancak veri kümesinin bir kısmı için iyi çalışır. Böylece, her model aslında topluluğun performansını artırıyor."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Bagging ve Boosting Kullanan Algoritmalar\n",
    "\n",
    "Bagging ve Boosting, makine öğreniminde en sık kullanılan tekniklerden ikisidir. Bu bölümde onlara detaylı olarak bakacağız. Odaklanacağımız algoritmalar şunlardır:\n",
    "\n",
    "Bagging algoritmaları:\n",
    "- Bagging meta-tahmincisi\n",
    "- Random Forest\n",
    "\n",
    "Boosting algoritmaları:\n",
    "\n",
    "- AdaBoost\n",
    "- GBM\n",
    "- XGBoost\n",
    "- Light GBM\n",
    "- CatBoost"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bagging meta-estimator"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "bagging meta-estimator, hem sınıflandırma hem de regresyon problemleri için kullanılabilen bir bagging algoritmasıdır. Tahminlerde bulunmak için tipik bagging tekniğini izler. bagging meta tahmin algoritması için adımlar şunlardır:\n",
    "\n",
    "1. Orijinal veri kümesinden (Bootstrapping) rastgele alt kümeler oluşturulur.\n",
    "2. Veri kümesinin alt kümesi tüm özellikleri içerir.\n",
    "3. Bu küçük setlerin her birine kullanıcı tarafından belirlenen bir temel tahmin aracı tanımlanmıştır.\n",
    "4. Nihai sonucu elde etmek için her modelden gelen tahminler birleştirilir."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Kod:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import cross_val_predict,cross_val_score\n",
    "from sklearn.metrics import confusion_matrix,accuracy_score,classification_report\n",
    "def print_score(clf, X_train, y_train, X_test, y_test):\n",
    "\n",
    "        print(\"Test Sonuçları:\\n\")        \n",
    "        print(\"accuracy score: {0:.4f}\\n\".format(accuracy_score(y_test, clf.predict(X_test))))\n",
    "        print(\"Classification Report: \\n {}\\n\".format(classification_report(y_test, clf.predict(X_test))))\n",
    "        print(\"Confusion Matrix: \\n {}\\n\".format(confusion_matrix(y_test, clf.predict(X_test))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Sonuçları:\n",
      "\n",
      "accuracy score: 0.7542\n",
      "\n",
      "Classification Report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.76      0.86      0.81       106\n",
      "           1       0.75      0.60      0.67        73\n",
      "\n",
      "    accuracy                           0.75       179\n",
      "   macro avg       0.75      0.73      0.74       179\n",
      "weighted avg       0.75      0.75      0.75       179\n",
      "\n",
      "\n",
      "Confusion Matrix: \n",
      " [[91 15]\n",
      " [29 44]]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import BaggingClassifier\n",
    "from sklearn import tree\n",
    "model = BaggingClassifier(tree.DecisionTreeClassifier(random_state=1))\n",
    "model.fit(X_train, y_train)\n",
    "acc_bag = model.score(X_test,y_test)\n",
    "print_score(model,X_train, y_train, X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Random Forest"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Random Forest, bagging tekniğini izleyen başka bir toplu makine öğrenme algoritmasıdır. bagging tahmin algoritmasının bir uzantısıdır. Rastgele ormandaki temel tahmin ediciler karar ağaçlarıdır. Torbalama meta tahmin edicisinin aksine, rastgele orman, karar ağacının her bir düğümünde en iyi bölünmeye karar vermek için kullanılan bir dizi özelliği rastgele seçer.\n",
    "\n",
    "Adım adım bakıldığında, rastgele bir orman modelinin yaptığı şey şunlardır:\n",
    "\n",
    "1. Orijinal veri kümesinden rastgele alt kümeler oluşturulur (önyükleme).\n",
    "2. Karar ağacındaki her düğümde, en iyi bölünmeye karar vermek için yalnızca rastgele bir özellik kümesi dikkate alınır.\n",
    "3. Her alt kümeye bir karar ağacı modeli uygulanır.\n",
    "4. Nihai tahmin, tüm karar ağaçlarından tahminlerin ortalaması alınarak hesaplanır.\n",
    "\n",
    "Özetlemek gerekirse, Rastgele Orman, veri noktalarını ve özelliklerini rastgele seçer ve birden çok ağaç oluşturur."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Kod:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Sonuçları:\n",
      "\n",
      "accuracy score: 0.7933\n",
      "\n",
      "Classification Report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.79      0.88      0.83       106\n",
      "           1       0.79      0.67      0.73        73\n",
      "\n",
      "    accuracy                           0.79       179\n",
      "   macro avg       0.79      0.77      0.78       179\n",
      "weighted avg       0.79      0.79      0.79       179\n",
      "\n",
      "\n",
      "Confusion Matrix: \n",
      " [[93 13]\n",
      " [24 49]]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import pandas as pd\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import accuracy_score\n",
    "model = RandomForestClassifier()\n",
    "model.fit(X_train,y_train)\n",
    "acc_rf = model.score(X_test,y_test)\n",
    "print_score(model,X_train, y_train, X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "importances = pd.DataFrame({'feature':X_train.columns,'importance':np.round(model.feature_importances_,3)})\n",
    "importances = importances.sort_values('importance',ascending=False).set_index('feature')\n",
    "importances.plot.bar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gradient Boosting\n",
    "\n",
    "Gradient Boosting veya GBM, hem regresyon hem de sınıflandırma problemleri için çalışan başka bir toplu makine öğrenme algoritmasıdır. GBM, güçlü bir öğrenci oluşturmak için bir dizi zayıf öğrenciyi birleştiren boosting tekniğini kullanır. Temel öğrenci olarak kullanılan regresyon ağaçları, ardışık her ağaç, bir önceki ağaç tarafından hesaplanan hatalar üzerine inşa edilir.\n",
    "\n",
    "\n",
    "GBM algoritmasını anlamak için basit bir örnek kullanacağız. Aşağıdaki verileri kullanarak bir grup insanın yaşını tahmin edelim:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](gbm1.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Ortalama yaş , veri kümesindeki tüm gözlemler için tahmin edilen değer olduğu varsayılır.\n",
    "2. Hatalar, bu ortalama yaş ve gerçek yaş değerleri kullanılarak hesaplanır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](gbm2.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3. Hedef değişken olarak yukarıda hesaplanan hatalar kullanılarak bir ağaç modeli oluşturulur. Amacımız, hatayı en aza indirmek için en iyi ayrımı bulmaktır.\n",
    "\n",
    "4. Bu modelin tahminleri prediction 1 sütununda toplanmıştır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](gbm3.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "5. Yukarıda hesaplanan bu değer yeni tahmindir.\n",
    "6. Bu tahmin edilen değer ve gerçek değer kullanılarak yeni hatalar hesaplanır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "![title](gbm4.png)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "7. Maksimum yineleme sayısına ulaşılıncaya kadar (veya hata işlevi değişmeyene) kadar 2'den 6'ya kadar olan adımlar tekrarlanır."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Kod:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Sonuçları:\n",
      "\n",
      "accuracy score: 0.8324\n",
      "\n",
      "Classification Report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.82      0.92      0.87       106\n",
      "           1       0.85      0.71      0.78        73\n",
      "\n",
      "    accuracy                           0.83       179\n",
      "   macro avg       0.84      0.81      0.82       179\n",
      "weighted avg       0.83      0.83      0.83       179\n",
      "\n",
      "\n",
      "Confusion Matrix: \n",
      " [[97  9]\n",
      " [21 52]]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import GradientBoostingClassifier\n",
    "model= GradientBoostingClassifier()\n",
    "model.fit(X_train, y_train)\n",
    "acc_gbm = model.score(X_test,y_test)\n",
    "print_score(model,X_train, y_train, X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "importances = pd.DataFrame({'feature':X_train.columns,'importance':np.round(model.feature_importances_,3)})\n",
    "importances = importances.sort_values('importance',ascending=False).set_index('feature')\n",
    "importances.plot.bar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## XGBoost"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "XGBoost (Extreme Gradient Boosting), gradient boosting algoritmasının gelişmiş bir uygulamasıdır. XGBoost, makine öğrenimi yarışmalarında ve hackathonlarda yaygın olarak kullanılan oldukça etkili bir makine öğrenimi algoritmasıdır. XGBoost, yüksek tahmin gücüne sahiptir ve diğer gradient boosting tekniklerinden neredeyse 10 kat daha hızlıdır. Ayrıca, yüksek varyansı azaltan ve genel performansı artıran çeşitli düzenlemeler içerir. Bu nedenle, \"regularized boosting\" tekniği olarak da bilinir.\n",
    "\n",
    "XGBoost'un diğer tekniklerden neden nispeten daha iyi olduğuna bakalım:\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### 1. Düzenlilik\n",
    "\n",
    "- Standart GBM uygulamasının XGBoost gibi bir düzenlenmesi yoktur.\n",
    "- Böylelikle XGBoost, aşırı öğrenmeyi azaltmaya da yardımcı olur.\n",
    "\n",
    "##### 2. Paralel İşleme\n",
    "\n",
    "- XGBoost paralel işlemeyi uygular ve GBM'den daha hızlıdır.\n",
    "- XGBoost ayrıca Hadoop'ta uygulamayı destekler.\n",
    "\n",
    "##### 3. Yüksek Eskenlik\n",
    "\n",
    "- XGBoost, kullanıcıların modele yepyeni bir boyut ekleyerek özel optimizasyon hedeflerini ve değerlendirme kriterlerini tanımlamalarına olanak tanır.\n",
    "\n",
    "##### 4. Eksik Değerleri Ele Alma\n",
    "\n",
    "- XGBoost, eksik değerlerin üstesinden gelmek için yerleşik özelliğe sahiptir.\n",
    "\n",
    "##### 5. Ağaç Budama\n",
    "\n",
    "- XGBoost, belirtilen maksimum derinliğe kadar bölmeler yapar ve ardından ağacı geriye doğru budamaya başlar ve ötesinde pozitif kazanım olmayan bölmeleri kaldırır.\n",
    "\n",
    "##### 6. Yerleşik Çapraz Doğrulama (Cross-Validation)\n",
    "\n",
    "- XGBoost, bir kullanıcının artırma işleminin her yinelemesinde bir çapraz doğrulama çalıştırmasına olanak tanır ve böylece tek bir çalıştırmada tam olarak optimum sayıda yükseltme yinelemesi elde etmek kolaydır.\n",
    "\n",
    "### Kod:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Sonuçları:\n",
      "\n",
      "accuracy score: 0.8101\n",
      "\n",
      "Classification Report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.80      0.92      0.85       106\n",
      "           1       0.84      0.66      0.74        73\n",
      "\n",
      "    accuracy                           0.81       179\n",
      "   macro avg       0.82      0.79      0.79       179\n",
      "weighted avg       0.81      0.81      0.81       179\n",
      "\n",
      "\n",
      "Confusion Matrix: \n",
      " [[97  9]\n",
      " [25 48]]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import xgboost as xgb\n",
    "model=xgb.XGBClassifier(learning_rate=0.1,subsample=0.3,min_child_weight=2,max_depth=4)\n",
    "model.fit(X_train, y_train)\n",
    "acc_xgb = model.score(X_test,y_test)\n",
    "print_score(model,X_train, y_train, X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "importances = pd.DataFrame({'feature':X_train.columns,'importance':np.round(model.feature_importances_,3)})\n",
    "importances = importances.sort_values('importance',ascending=False).set_index('feature')\n",
    "importances.plot.bar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## LightGBM\n",
    "\n",
    "Light GBM'nin nasıl çalıştığını tartışmadan önce, bu kadar çok algoritmaya sahip olduğumuz halde neden bu algoritmaya ihtiyacımız olduğunuz anlayalım. Light GBM, veri kümesi çok büyük olduğunda en çok tercih edilen algoritmadır. Diğer algoritmalarla karşılaştırıldığında, Light GBM'in büyük bir veri kümesinde çalışması daha az zaman alır."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Sonuçları:\n",
      "\n",
      "accuracy score: 0.7877\n",
      "\n",
      "Classification Report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.77      0.91      0.83       106\n",
      "           1       0.82      0.62      0.70        73\n",
      "\n",
      "    accuracy                           0.79       179\n",
      "   macro avg       0.80      0.76      0.77       179\n",
      "weighted avg       0.79      0.79      0.78       179\n",
      "\n",
      "\n",
      "Confusion Matrix: \n",
      " [[96 10]\n",
      " [28 45]]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from lightgbm import LGBMClassifier\n",
    "model = LGBMClassifier()\n",
    "model.fit(X_train,y_train)\n",
    "acc_lgb = model.score(X_test,y_test)\n",
    "print_score(model,X_train, y_train, X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "importances = pd.DataFrame({'feature':X_train.columns,'importance':np.round(model.feature_importances_,3)})\n",
    "importances = importances.sort_values('importance',ascending=False).set_index('feature')\n",
    "importances.plot.bar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## CatBoost\n",
    "\n",
    "Kategorik değişkenleri ele almak, özellikle çok sayıda bu tür değişkenlere sahip olduğunuzda sıkıcı bir süreçtir. Kategorik değişkenleriniz çok fazla etikete sahip olduğunda (yani bunlar çok önemliyse), üzerlerinde one-hot-encoding yapmak katlanarak boyutluluğu artırır ve veri kümesiyle çalışmak gerçekten zor hale gelir.\n",
    "\n",
    "CatBoost, kategorik değişkenlerle otomatik olarak ilgilenebilir ve diğer makine öğrenimi algoritmaları gibi kapsamlı veri ön işlemesi gerektirmez. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Learning rate set to 0.007881\n",
      "0:\tlearn: 0.6869339\ttotal: 179ms\tremaining: 2m 58s\n",
      "1:\tlearn: 0.6814352\ttotal: 180ms\tremaining: 1m 30s\n",
      "2:\tlearn: 0.6763111\ttotal: 182ms\tremaining: 1m\n",
      "3:\tlearn: 0.6712485\ttotal: 183ms\tremaining: 45.6s\n",
      "4:\tlearn: 0.6677466\ttotal: 184ms\tremaining: 36.6s\n",
      "5:\tlearn: 0.6649968\ttotal: 185ms\tremaining: 30.6s\n",
      "6:\tlearn: 0.6591619\ttotal: 186ms\tremaining: 26.4s\n",
      "7:\tlearn: 0.6541660\ttotal: 187ms\tremaining: 23.2s\n",
      "8:\tlearn: 0.6501075\ttotal: 188ms\tremaining: 20.7s\n",
      "9:\tlearn: 0.6450524\ttotal: 190ms\tremaining: 18.8s\n",
      "10:\tlearn: 0.6396774\ttotal: 191ms\tremaining: 17.2s\n",
      "11:\tlearn: 0.6343361\ttotal: 192ms\tremaining: 15.8s\n",
      "12:\tlearn: 0.6291343\ttotal: 194ms\tremaining: 14.7s\n",
      "13:\tlearn: 0.6250985\ttotal: 195ms\tremaining: 13.8s\n",
      "14:\tlearn: 0.6212028\ttotal: 197ms\tremaining: 12.9s\n",
      "15:\tlearn: 0.6170359\ttotal: 198ms\tremaining: 12.2s\n",
      "16:\tlearn: 0.6120818\ttotal: 199ms\tremaining: 11.5s\n",
      "17:\tlearn: 0.6074254\ttotal: 201ms\tremaining: 10.9s\n",
      "18:\tlearn: 0.6042157\ttotal: 202ms\tremaining: 10.4s\n",
      "19:\tlearn: 0.6010588\ttotal: 202ms\tremaining: 9.91s\n",
      "20:\tlearn: 0.5965770\ttotal: 204ms\tremaining: 9.49s\n",
      "21:\tlearn: 0.5929426\ttotal: 205ms\tremaining: 9.11s\n",
      "22:\tlearn: 0.5893689\ttotal: 206ms\tremaining: 8.74s\n",
      "23:\tlearn: 0.5871231\ttotal: 206ms\tremaining: 8.39s\n",
      "24:\tlearn: 0.5831945\ttotal: 208ms\tremaining: 8.1s\n",
      "25:\tlearn: 0.5796250\ttotal: 209ms\tremaining: 7.82s\n",
      "26:\tlearn: 0.5757451\ttotal: 210ms\tremaining: 7.56s\n",
      "27:\tlearn: 0.5718254\ttotal: 211ms\tremaining: 7.33s\n",
      "28:\tlearn: 0.5688318\ttotal: 212ms\tremaining: 7.11s\n",
      "29:\tlearn: 0.5646421\ttotal: 213ms\tremaining: 6.9s\n",
      "30:\tlearn: 0.5610623\ttotal: 215ms\tremaining: 6.71s\n",
      "31:\tlearn: 0.5582025\ttotal: 216ms\tremaining: 6.53s\n",
      "32:\tlearn: 0.5550707\ttotal: 218ms\tremaining: 6.38s\n",
      "33:\tlearn: 0.5517542\ttotal: 219ms\tremaining: 6.22s\n",
      "34:\tlearn: 0.5481118\ttotal: 221ms\tremaining: 6.09s\n",
      "35:\tlearn: 0.5448491\ttotal: 222ms\tremaining: 5.96s\n",
      "36:\tlearn: 0.5412271\ttotal: 223ms\tremaining: 5.81s\n",
      "37:\tlearn: 0.5379209\ttotal: 225ms\tremaining: 5.68s\n",
      "38:\tlearn: 0.5346580\ttotal: 226ms\tremaining: 5.56s\n",
      "39:\tlearn: 0.5313487\ttotal: 227ms\tremaining: 5.44s\n",
      "40:\tlearn: 0.5281063\ttotal: 228ms\tremaining: 5.32s\n",
      "41:\tlearn: 0.5256447\ttotal: 229ms\tremaining: 5.22s\n",
      "42:\tlearn: 0.5225151\ttotal: 230ms\tremaining: 5.12s\n",
      "43:\tlearn: 0.5197667\ttotal: 231ms\tremaining: 5.02s\n",
      "44:\tlearn: 0.5178806\ttotal: 232ms\tremaining: 4.93s\n",
      "45:\tlearn: 0.5149179\ttotal: 233ms\tremaining: 4.84s\n",
      "46:\tlearn: 0.5118936\ttotal: 234ms\tremaining: 4.75s\n",
      "47:\tlearn: 0.5095418\ttotal: 235ms\tremaining: 4.67s\n",
      "48:\tlearn: 0.5069018\ttotal: 237ms\tremaining: 4.59s\n",
      "49:\tlearn: 0.5045027\ttotal: 238ms\tremaining: 4.51s\n",
      "50:\tlearn: 0.5022078\ttotal: 239ms\tremaining: 4.44s\n",
      "51:\tlearn: 0.5009028\ttotal: 240ms\tremaining: 4.37s\n",
      "52:\tlearn: 0.4988188\ttotal: 241ms\tremaining: 4.3s\n",
      "53:\tlearn: 0.4967572\ttotal: 242ms\tremaining: 4.24s\n",
      "54:\tlearn: 0.4947700\ttotal: 243ms\tremaining: 4.17s\n",
      "55:\tlearn: 0.4925983\ttotal: 245ms\tremaining: 4.12s\n",
      "56:\tlearn: 0.4904389\ttotal: 246ms\tremaining: 4.07s\n",
      "57:\tlearn: 0.4878437\ttotal: 247ms\tremaining: 4.02s\n",
      "58:\tlearn: 0.4856584\ttotal: 248ms\tremaining: 3.96s\n",
      "59:\tlearn: 0.4840084\ttotal: 249ms\tremaining: 3.91s\n",
      "60:\tlearn: 0.4818933\ttotal: 251ms\tremaining: 3.86s\n",
      "61:\tlearn: 0.4795636\ttotal: 252ms\tremaining: 3.81s\n",
      "62:\tlearn: 0.4777653\ttotal: 253ms\tremaining: 3.76s\n",
      "63:\tlearn: 0.4756889\ttotal: 254ms\tremaining: 3.71s\n",
      "64:\tlearn: 0.4741442\ttotal: 255ms\tremaining: 3.67s\n",
      "65:\tlearn: 0.4724859\ttotal: 256ms\tremaining: 3.63s\n",
      "66:\tlearn: 0.4703561\ttotal: 257ms\tremaining: 3.58s\n",
      "67:\tlearn: 0.4686564\ttotal: 259ms\tremaining: 3.54s\n",
      "68:\tlearn: 0.4668923\ttotal: 260ms\tremaining: 3.5s\n",
      "69:\tlearn: 0.4652055\ttotal: 261ms\tremaining: 3.46s\n",
      "70:\tlearn: 0.4634507\ttotal: 262ms\tremaining: 3.42s\n",
      "71:\tlearn: 0.4616975\ttotal: 263ms\tremaining: 3.38s\n",
      "72:\tlearn: 0.4600555\ttotal: 264ms\tremaining: 3.35s\n",
      "73:\tlearn: 0.4587947\ttotal: 265ms\tremaining: 3.31s\n",
      "74:\tlearn: 0.4569169\ttotal: 266ms\tremaining: 3.28s\n",
      "75:\tlearn: 0.4552132\ttotal: 267ms\tremaining: 3.25s\n",
      "76:\tlearn: 0.4535425\ttotal: 268ms\tremaining: 3.21s\n",
      "77:\tlearn: 0.4518319\ttotal: 269ms\tremaining: 3.18s\n",
      "78:\tlearn: 0.4503408\ttotal: 270ms\tremaining: 3.15s\n",
      "79:\tlearn: 0.4487482\ttotal: 272ms\tremaining: 3.12s\n",
      "80:\tlearn: 0.4471483\ttotal: 273ms\tremaining: 3.09s\n",
      "81:\tlearn: 0.4455122\ttotal: 274ms\tremaining: 3.07s\n",
      "82:\tlearn: 0.4442221\ttotal: 275ms\tremaining: 3.04s\n",
      "83:\tlearn: 0.4426689\ttotal: 276ms\tremaining: 3.01s\n",
      "84:\tlearn: 0.4411585\ttotal: 277ms\tremaining: 2.98s\n",
      "85:\tlearn: 0.4396209\ttotal: 278ms\tremaining: 2.96s\n",
      "86:\tlearn: 0.4385386\ttotal: 279ms\tremaining: 2.93s\n",
      "87:\tlearn: 0.4370425\ttotal: 280ms\tremaining: 2.9s\n",
      "88:\tlearn: 0.4355059\ttotal: 281ms\tremaining: 2.88s\n",
      "89:\tlearn: 0.4339610\ttotal: 282ms\tremaining: 2.85s\n",
      "90:\tlearn: 0.4326355\ttotal: 283ms\tremaining: 2.83s\n",
      "91:\tlearn: 0.4314599\ttotal: 284ms\tremaining: 2.81s\n",
      "92:\tlearn: 0.4305998\ttotal: 285ms\tremaining: 2.78s\n",
      "93:\tlearn: 0.4293935\ttotal: 287ms\tremaining: 2.76s\n",
      "94:\tlearn: 0.4285209\ttotal: 287ms\tremaining: 2.74s\n",
      "95:\tlearn: 0.4275796\ttotal: 288ms\tremaining: 2.71s\n",
      "96:\tlearn: 0.4265343\ttotal: 289ms\tremaining: 2.69s\n",
      "97:\tlearn: 0.4250575\ttotal: 290ms\tremaining: 2.67s\n",
      "98:\tlearn: 0.4236656\ttotal: 292ms\tremaining: 2.65s\n",
      "99:\tlearn: 0.4225113\ttotal: 293ms\tremaining: 2.63s\n",
      "100:\tlearn: 0.4215123\ttotal: 294ms\tremaining: 2.61s\n",
      "101:\tlearn: 0.4203080\ttotal: 295ms\tremaining: 2.6s\n",
      "102:\tlearn: 0.4192462\ttotal: 296ms\tremaining: 2.58s\n",
      "103:\tlearn: 0.4183070\ttotal: 297ms\tremaining: 2.56s\n",
      "104:\tlearn: 0.4171065\ttotal: 298ms\tremaining: 2.54s\n",
      "105:\tlearn: 0.4161865\ttotal: 299ms\tremaining: 2.52s\n",
      "106:\tlearn: 0.4150760\ttotal: 300ms\tremaining: 2.51s\n",
      "107:\tlearn: 0.4143477\ttotal: 301ms\tremaining: 2.49s\n",
      "108:\tlearn: 0.4135538\ttotal: 303ms\tremaining: 2.47s\n",
      "109:\tlearn: 0.4123604\ttotal: 304ms\tremaining: 2.46s\n",
      "110:\tlearn: 0.4113797\ttotal: 305ms\tremaining: 2.44s\n",
      "111:\tlearn: 0.4107620\ttotal: 306ms\tremaining: 2.43s\n",
      "112:\tlearn: 0.4100270\ttotal: 307ms\tremaining: 2.41s\n",
      "113:\tlearn: 0.4092766\ttotal: 308ms\tremaining: 2.4s\n",
      "114:\tlearn: 0.4081929\ttotal: 309ms\tremaining: 2.38s\n",
      "115:\tlearn: 0.4073196\ttotal: 310ms\tremaining: 2.37s\n",
      "116:\tlearn: 0.4066614\ttotal: 312ms\tremaining: 2.35s\n",
      "117:\tlearn: 0.4059519\ttotal: 313ms\tremaining: 2.34s\n",
      "118:\tlearn: 0.4050161\ttotal: 314ms\tremaining: 2.33s\n",
      "119:\tlearn: 0.4041277\ttotal: 316ms\tremaining: 2.32s\n",
      "120:\tlearn: 0.4032449\ttotal: 318ms\tremaining: 2.31s\n",
      "121:\tlearn: 0.4024481\ttotal: 319ms\tremaining: 2.29s\n",
      "122:\tlearn: 0.4018155\ttotal: 320ms\tremaining: 2.28s\n",
      "123:\tlearn: 0.4011716\ttotal: 321ms\tremaining: 2.27s\n",
      "124:\tlearn: 0.4004074\ttotal: 322ms\tremaining: 2.25s\n",
      "125:\tlearn: 0.4001237\ttotal: 322ms\tremaining: 2.24s\n",
      "126:\tlearn: 0.3993852\ttotal: 324ms\tremaining: 2.22s\n",
      "127:\tlearn: 0.3988580\ttotal: 325ms\tremaining: 2.21s\n",
      "128:\tlearn: 0.3982154\ttotal: 326ms\tremaining: 2.2s\n",
      "129:\tlearn: 0.3973786\ttotal: 327ms\tremaining: 2.19s\n",
      "130:\tlearn: 0.3970035\ttotal: 328ms\tremaining: 2.17s\n",
      "131:\tlearn: 0.3962012\ttotal: 329ms\tremaining: 2.16s\n",
      "132:\tlearn: 0.3955072\ttotal: 330ms\tremaining: 2.15s\n",
      "133:\tlearn: 0.3949940\ttotal: 331ms\tremaining: 2.14s\n",
      "134:\tlearn: 0.3941520\ttotal: 332ms\tremaining: 2.13s\n",
      "135:\tlearn: 0.3933775\ttotal: 333ms\tremaining: 2.12s\n",
      "136:\tlearn: 0.3927144\ttotal: 334ms\tremaining: 2.1s\n",
      "137:\tlearn: 0.3918673\ttotal: 335ms\tremaining: 2.09s\n",
      "138:\tlearn: 0.3912247\ttotal: 336ms\tremaining: 2.08s\n",
      "139:\tlearn: 0.3904337\ttotal: 337ms\tremaining: 2.07s\n",
      "140:\tlearn: 0.3898549\ttotal: 339ms\tremaining: 2.06s\n",
      "141:\tlearn: 0.3891117\ttotal: 340ms\tremaining: 2.05s\n",
      "142:\tlearn: 0.3882961\ttotal: 341ms\tremaining: 2.04s\n",
      "143:\tlearn: 0.3878330\ttotal: 342ms\tremaining: 2.03s\n",
      "144:\tlearn: 0.3872950\ttotal: 344ms\tremaining: 2.02s\n",
      "145:\tlearn: 0.3866448\ttotal: 345ms\tremaining: 2.02s\n",
      "146:\tlearn: 0.3861911\ttotal: 346ms\tremaining: 2.01s\n",
      "147:\tlearn: 0.3855133\ttotal: 347ms\tremaining: 2s\n",
      "148:\tlearn: 0.3852571\ttotal: 348ms\tremaining: 1.99s\n",
      "149:\tlearn: 0.3846193\ttotal: 349ms\tremaining: 1.98s\n",
      "150:\tlearn: 0.3838456\ttotal: 350ms\tremaining: 1.97s\n",
      "151:\tlearn: 0.3832336\ttotal: 351ms\tremaining: 1.96s\n",
      "152:\tlearn: 0.3825697\ttotal: 352ms\tremaining: 1.95s\n",
      "153:\tlearn: 0.3818166\ttotal: 353ms\tremaining: 1.94s\n",
      "154:\tlearn: 0.3812147\ttotal: 354ms\tremaining: 1.93s\n",
      "155:\tlearn: 0.3805763\ttotal: 356ms\tremaining: 1.92s\n",
      "156:\tlearn: 0.3800553\ttotal: 357ms\tremaining: 1.92s\n",
      "157:\tlearn: 0.3795512\ttotal: 359ms\tremaining: 1.91s\n",
      "158:\tlearn: 0.3791030\ttotal: 360ms\tremaining: 1.9s\n",
      "159:\tlearn: 0.3785062\ttotal: 361ms\tremaining: 1.9s\n",
      "160:\tlearn: 0.3780343\ttotal: 362ms\tremaining: 1.89s\n",
      "161:\tlearn: 0.3775572\ttotal: 363ms\tremaining: 1.88s\n",
      "162:\tlearn: 0.3768152\ttotal: 365ms\tremaining: 1.87s\n",
      "163:\tlearn: 0.3765495\ttotal: 366ms\tremaining: 1.87s\n",
      "164:\tlearn: 0.3760320\ttotal: 368ms\tremaining: 1.86s\n",
      "165:\tlearn: 0.3754397\ttotal: 369ms\tremaining: 1.85s\n",
      "166:\tlearn: 0.3750876\ttotal: 370ms\tremaining: 1.85s\n",
      "167:\tlearn: 0.3746156\ttotal: 371ms\tremaining: 1.84s\n",
      "168:\tlearn: 0.3742321\ttotal: 372ms\tremaining: 1.83s\n",
      "169:\tlearn: 0.3737489\ttotal: 373ms\tremaining: 1.82s\n",
      "170:\tlearn: 0.3733981\ttotal: 375ms\tremaining: 1.81s\n",
      "171:\tlearn: 0.3729643\ttotal: 376ms\tremaining: 1.81s\n",
      "172:\tlearn: 0.3725815\ttotal: 377ms\tremaining: 1.8s\n",
      "173:\tlearn: 0.3722019\ttotal: 378ms\tremaining: 1.79s\n",
      "174:\tlearn: 0.3717358\ttotal: 379ms\tremaining: 1.79s\n",
      "175:\tlearn: 0.3710891\ttotal: 380ms\tremaining: 1.78s\n",
      "176:\tlearn: 0.3704972\ttotal: 381ms\tremaining: 1.77s\n",
      "177:\tlearn: 0.3699054\ttotal: 382ms\tremaining: 1.76s\n",
      "178:\tlearn: 0.3695202\ttotal: 383ms\tremaining: 1.76s\n",
      "179:\tlearn: 0.3692268\ttotal: 385ms\tremaining: 1.75s\n",
      "180:\tlearn: 0.3687254\ttotal: 386ms\tremaining: 1.75s\n",
      "181:\tlearn: 0.3683551\ttotal: 387ms\tremaining: 1.74s\n",
      "182:\tlearn: 0.3677812\ttotal: 388ms\tremaining: 1.73s\n",
      "183:\tlearn: 0.3673412\ttotal: 389ms\tremaining: 1.72s\n",
      "184:\tlearn: 0.3670165\ttotal: 390ms\tremaining: 1.72s\n",
      "185:\tlearn: 0.3666985\ttotal: 391ms\tremaining: 1.71s\n",
      "186:\tlearn: 0.3662679\ttotal: 392ms\tremaining: 1.71s\n",
      "187:\tlearn: 0.3658816\ttotal: 393ms\tremaining: 1.7s\n",
      "188:\tlearn: 0.3656895\ttotal: 394ms\tremaining: 1.69s\n",
      "189:\tlearn: 0.3652806\ttotal: 395ms\tremaining: 1.68s\n",
      "190:\tlearn: 0.3648226\ttotal: 397ms\tremaining: 1.68s\n",
      "191:\tlearn: 0.3643485\ttotal: 398ms\tremaining: 1.67s\n",
      "192:\tlearn: 0.3640834\ttotal: 399ms\tremaining: 1.67s\n",
      "193:\tlearn: 0.3637350\ttotal: 400ms\tremaining: 1.66s\n",
      "194:\tlearn: 0.3634719\ttotal: 401ms\tremaining: 1.65s\n",
      "195:\tlearn: 0.3631167\ttotal: 402ms\tremaining: 1.65s\n",
      "196:\tlearn: 0.3627713\ttotal: 403ms\tremaining: 1.64s\n",
      "197:\tlearn: 0.3624498\ttotal: 404ms\tremaining: 1.64s\n",
      "198:\tlearn: 0.3620806\ttotal: 405ms\tremaining: 1.63s\n",
      "199:\tlearn: 0.3619591\ttotal: 406ms\tremaining: 1.62s\n",
      "200:\tlearn: 0.3616635\ttotal: 407ms\tremaining: 1.62s\n",
      "201:\tlearn: 0.3613746\ttotal: 408ms\tremaining: 1.61s\n",
      "202:\tlearn: 0.3610168\ttotal: 409ms\tremaining: 1.6s\n",
      "203:\tlearn: 0.3607565\ttotal: 410ms\tremaining: 1.6s\n",
      "204:\tlearn: 0.3604901\ttotal: 411ms\tremaining: 1.59s\n",
      "205:\tlearn: 0.3601104\ttotal: 412ms\tremaining: 1.59s\n",
      "206:\tlearn: 0.3596381\ttotal: 413ms\tremaining: 1.58s\n",
      "207:\tlearn: 0.3592285\ttotal: 414ms\tremaining: 1.57s\n",
      "208:\tlearn: 0.3590141\ttotal: 415ms\tremaining: 1.57s\n",
      "209:\tlearn: 0.3587560\ttotal: 416ms\tremaining: 1.56s\n",
      "210:\tlearn: 0.3584837\ttotal: 417ms\tremaining: 1.56s\n",
      "211:\tlearn: 0.3582461\ttotal: 418ms\tremaining: 1.55s\n",
      "212:\tlearn: 0.3580036\ttotal: 419ms\tremaining: 1.55s\n",
      "213:\tlearn: 0.3578621\ttotal: 419ms\tremaining: 1.54s\n",
      "214:\tlearn: 0.3573093\ttotal: 420ms\tremaining: 1.53s\n",
      "215:\tlearn: 0.3569921\ttotal: 421ms\tremaining: 1.53s\n",
      "216:\tlearn: 0.3568552\ttotal: 422ms\tremaining: 1.52s\n",
      "217:\tlearn: 0.3565331\ttotal: 423ms\tremaining: 1.52s\n",
      "218:\tlearn: 0.3562792\ttotal: 424ms\tremaining: 1.51s\n",
      "219:\tlearn: 0.3560034\ttotal: 425ms\tremaining: 1.51s\n",
      "220:\tlearn: 0.3558593\ttotal: 426ms\tremaining: 1.5s\n",
      "221:\tlearn: 0.3557059\ttotal: 427ms\tremaining: 1.5s\n",
      "222:\tlearn: 0.3552749\ttotal: 428ms\tremaining: 1.49s\n",
      "223:\tlearn: 0.3550063\ttotal: 429ms\tremaining: 1.49s\n",
      "224:\tlearn: 0.3546820\ttotal: 430ms\tremaining: 1.48s\n",
      "225:\tlearn: 0.3544775\ttotal: 431ms\tremaining: 1.48s\n",
      "226:\tlearn: 0.3542062\ttotal: 432ms\tremaining: 1.47s\n",
      "227:\tlearn: 0.3540699\ttotal: 433ms\tremaining: 1.47s\n",
      "228:\tlearn: 0.3538681\ttotal: 434ms\tremaining: 1.46s\n",
      "229:\tlearn: 0.3535639\ttotal: 436ms\tremaining: 1.46s\n",
      "230:\tlearn: 0.3533398\ttotal: 437ms\tremaining: 1.45s\n",
      "231:\tlearn: 0.3531574\ttotal: 438ms\tremaining: 1.45s\n",
      "232:\tlearn: 0.3528169\ttotal: 439ms\tremaining: 1.44s\n",
      "233:\tlearn: 0.3525161\ttotal: 440ms\tremaining: 1.44s\n",
      "234:\tlearn: 0.3521227\ttotal: 441ms\tremaining: 1.43s\n",
      "235:\tlearn: 0.3520244\ttotal: 442ms\tremaining: 1.43s\n",
      "236:\tlearn: 0.3518180\ttotal: 443ms\tremaining: 1.43s\n",
      "237:\tlearn: 0.3513649\ttotal: 444ms\tremaining: 1.42s\n",
      "238:\tlearn: 0.3513268\ttotal: 444ms\tremaining: 1.41s\n",
      "239:\tlearn: 0.3512632\ttotal: 445ms\tremaining: 1.41s\n",
      "240:\tlearn: 0.3511679\ttotal: 446ms\tremaining: 1.4s\n",
      "241:\tlearn: 0.3508888\ttotal: 447ms\tremaining: 1.4s\n",
      "242:\tlearn: 0.3506095\ttotal: 448ms\tremaining: 1.4s\n",
      "243:\tlearn: 0.3503316\ttotal: 450ms\tremaining: 1.39s\n",
      "244:\tlearn: 0.3500132\ttotal: 451ms\tremaining: 1.39s\n",
      "245:\tlearn: 0.3497673\ttotal: 452ms\tremaining: 1.38s\n",
      "246:\tlearn: 0.3494810\ttotal: 453ms\tremaining: 1.38s\n",
      "247:\tlearn: 0.3491671\ttotal: 454ms\tremaining: 1.38s\n",
      "248:\tlearn: 0.3489636\ttotal: 455ms\tremaining: 1.37s\n",
      "249:\tlearn: 0.3487621\ttotal: 456ms\tremaining: 1.37s\n",
      "250:\tlearn: 0.3486700\ttotal: 456ms\tremaining: 1.36s\n",
      "251:\tlearn: 0.3484835\ttotal: 457ms\tremaining: 1.36s\n",
      "252:\tlearn: 0.3481784\ttotal: 458ms\tremaining: 1.35s\n",
      "253:\tlearn: 0.3477789\ttotal: 459ms\tremaining: 1.35s\n",
      "254:\tlearn: 0.3473848\ttotal: 460ms\tremaining: 1.34s\n",
      "255:\tlearn: 0.3470336\ttotal: 461ms\tremaining: 1.34s\n",
      "256:\tlearn: 0.3468033\ttotal: 462ms\tremaining: 1.34s\n",
      "257:\tlearn: 0.3466364\ttotal: 463ms\tremaining: 1.33s\n",
      "258:\tlearn: 0.3463243\ttotal: 464ms\tremaining: 1.33s\n",
      "259:\tlearn: 0.3460826\ttotal: 465ms\tremaining: 1.32s\n",
      "260:\tlearn: 0.3458162\ttotal: 467ms\tremaining: 1.32s\n",
      "261:\tlearn: 0.3455734\ttotal: 468ms\tremaining: 1.32s\n",
      "262:\tlearn: 0.3450312\ttotal: 469ms\tremaining: 1.31s\n",
      "263:\tlearn: 0.3449239\ttotal: 469ms\tremaining: 1.31s\n",
      "264:\tlearn: 0.3446477\ttotal: 470ms\tremaining: 1.3s\n",
      "265:\tlearn: 0.3443832\ttotal: 471ms\tremaining: 1.3s\n",
      "266:\tlearn: 0.3441396\ttotal: 472ms\tremaining: 1.3s\n",
      "267:\tlearn: 0.3438934\ttotal: 473ms\tremaining: 1.29s\n",
      "268:\tlearn: 0.3434717\ttotal: 474ms\tremaining: 1.29s\n",
      "269:\tlearn: 0.3433464\ttotal: 475ms\tremaining: 1.28s\n",
      "270:\tlearn: 0.3431058\ttotal: 476ms\tremaining: 1.28s\n",
      "271:\tlearn: 0.3428089\ttotal: 477ms\tremaining: 1.28s\n",
      "272:\tlearn: 0.3426734\ttotal: 478ms\tremaining: 1.27s\n",
      "273:\tlearn: 0.3422967\ttotal: 479ms\tremaining: 1.27s\n",
      "274:\tlearn: 0.3421198\ttotal: 480ms\tremaining: 1.27s\n",
      "275:\tlearn: 0.3418017\ttotal: 481ms\tremaining: 1.26s\n",
      "276:\tlearn: 0.3415970\ttotal: 483ms\tremaining: 1.26s\n",
      "277:\tlearn: 0.3412893\ttotal: 484ms\tremaining: 1.25s\n",
      "278:\tlearn: 0.3411784\ttotal: 484ms\tremaining: 1.25s\n",
      "279:\tlearn: 0.3410169\ttotal: 485ms\tremaining: 1.25s\n",
      "280:\tlearn: 0.3407581\ttotal: 486ms\tremaining: 1.24s\n",
      "281:\tlearn: 0.3405164\ttotal: 487ms\tremaining: 1.24s\n",
      "282:\tlearn: 0.3403374\ttotal: 488ms\tremaining: 1.24s\n",
      "283:\tlearn: 0.3401397\ttotal: 489ms\tremaining: 1.23s\n",
      "284:\tlearn: 0.3399940\ttotal: 490ms\tremaining: 1.23s\n",
      "285:\tlearn: 0.3398536\ttotal: 491ms\tremaining: 1.23s\n",
      "286:\tlearn: 0.3397258\ttotal: 492ms\tremaining: 1.22s\n",
      "287:\tlearn: 0.3394457\ttotal: 493ms\tremaining: 1.22s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "288:\tlearn: 0.3391078\ttotal: 495ms\tremaining: 1.22s\n",
      "289:\tlearn: 0.3389093\ttotal: 496ms\tremaining: 1.21s\n",
      "290:\tlearn: 0.3386576\ttotal: 497ms\tremaining: 1.21s\n",
      "291:\tlearn: 0.3385004\ttotal: 498ms\tremaining: 1.21s\n",
      "292:\tlearn: 0.3383693\ttotal: 499ms\tremaining: 1.2s\n",
      "293:\tlearn: 0.3381678\ttotal: 500ms\tremaining: 1.2s\n",
      "294:\tlearn: 0.3379514\ttotal: 501ms\tremaining: 1.2s\n",
      "295:\tlearn: 0.3375961\ttotal: 502ms\tremaining: 1.19s\n",
      "296:\tlearn: 0.3372951\ttotal: 503ms\tremaining: 1.19s\n",
      "297:\tlearn: 0.3371168\ttotal: 505ms\tremaining: 1.19s\n",
      "298:\tlearn: 0.3369695\ttotal: 506ms\tremaining: 1.19s\n",
      "299:\tlearn: 0.3367523\ttotal: 507ms\tremaining: 1.18s\n",
      "300:\tlearn: 0.3365992\ttotal: 509ms\tremaining: 1.18s\n",
      "301:\tlearn: 0.3364605\ttotal: 510ms\tremaining: 1.18s\n",
      "302:\tlearn: 0.3362488\ttotal: 511ms\tremaining: 1.18s\n",
      "303:\tlearn: 0.3360246\ttotal: 513ms\tremaining: 1.17s\n",
      "304:\tlearn: 0.3359261\ttotal: 514ms\tremaining: 1.17s\n",
      "305:\tlearn: 0.3357735\ttotal: 515ms\tremaining: 1.17s\n",
      "306:\tlearn: 0.3354919\ttotal: 516ms\tremaining: 1.16s\n",
      "307:\tlearn: 0.3351057\ttotal: 517ms\tremaining: 1.16s\n",
      "308:\tlearn: 0.3347536\ttotal: 518ms\tremaining: 1.16s\n",
      "309:\tlearn: 0.3344132\ttotal: 519ms\tremaining: 1.15s\n",
      "310:\tlearn: 0.3343123\ttotal: 520ms\tremaining: 1.15s\n",
      "311:\tlearn: 0.3342206\ttotal: 521ms\tremaining: 1.15s\n",
      "312:\tlearn: 0.3341032\ttotal: 522ms\tremaining: 1.14s\n",
      "313:\tlearn: 0.3339117\ttotal: 523ms\tremaining: 1.14s\n",
      "314:\tlearn: 0.3338096\ttotal: 523ms\tremaining: 1.14s\n",
      "315:\tlearn: 0.3335741\ttotal: 524ms\tremaining: 1.13s\n",
      "316:\tlearn: 0.3334631\ttotal: 525ms\tremaining: 1.13s\n",
      "317:\tlearn: 0.3333037\ttotal: 526ms\tremaining: 1.13s\n",
      "318:\tlearn: 0.3330311\ttotal: 527ms\tremaining: 1.13s\n",
      "319:\tlearn: 0.3328790\ttotal: 528ms\tremaining: 1.12s\n",
      "320:\tlearn: 0.3326885\ttotal: 529ms\tremaining: 1.12s\n",
      "321:\tlearn: 0.3325250\ttotal: 530ms\tremaining: 1.12s\n",
      "322:\tlearn: 0.3323578\ttotal: 531ms\tremaining: 1.11s\n",
      "323:\tlearn: 0.3321912\ttotal: 533ms\tremaining: 1.11s\n",
      "324:\tlearn: 0.3319253\ttotal: 534ms\tremaining: 1.11s\n",
      "325:\tlearn: 0.3316941\ttotal: 535ms\tremaining: 1.1s\n",
      "326:\tlearn: 0.3315329\ttotal: 536ms\tremaining: 1.1s\n",
      "327:\tlearn: 0.3313815\ttotal: 537ms\tremaining: 1.1s\n",
      "328:\tlearn: 0.3311155\ttotal: 538ms\tremaining: 1.1s\n",
      "329:\tlearn: 0.3308763\ttotal: 539ms\tremaining: 1.09s\n",
      "330:\tlearn: 0.3308026\ttotal: 540ms\tremaining: 1.09s\n",
      "331:\tlearn: 0.3306116\ttotal: 541ms\tremaining: 1.09s\n",
      "332:\tlearn: 0.3304102\ttotal: 542ms\tremaining: 1.08s\n",
      "333:\tlearn: 0.3302614\ttotal: 543ms\tremaining: 1.08s\n",
      "334:\tlearn: 0.3299201\ttotal: 544ms\tremaining: 1.08s\n",
      "335:\tlearn: 0.3297833\ttotal: 545ms\tremaining: 1.08s\n",
      "336:\tlearn: 0.3296652\ttotal: 546ms\tremaining: 1.07s\n",
      "337:\tlearn: 0.3295286\ttotal: 547ms\tremaining: 1.07s\n",
      "338:\tlearn: 0.3293642\ttotal: 550ms\tremaining: 1.07s\n",
      "339:\tlearn: 0.3292018\ttotal: 551ms\tremaining: 1.07s\n",
      "340:\tlearn: 0.3290346\ttotal: 552ms\tremaining: 1.07s\n",
      "341:\tlearn: 0.3287770\ttotal: 553ms\tremaining: 1.06s\n",
      "342:\tlearn: 0.3285992\ttotal: 554ms\tremaining: 1.06s\n",
      "343:\tlearn: 0.3284602\ttotal: 555ms\tremaining: 1.06s\n",
      "344:\tlearn: 0.3281305\ttotal: 556ms\tremaining: 1.06s\n",
      "345:\tlearn: 0.3281195\ttotal: 557ms\tremaining: 1.05s\n",
      "346:\tlearn: 0.3279763\ttotal: 558ms\tremaining: 1.05s\n",
      "347:\tlearn: 0.3278294\ttotal: 559ms\tremaining: 1.05s\n",
      "348:\tlearn: 0.3274527\ttotal: 560ms\tremaining: 1.04s\n",
      "349:\tlearn: 0.3271363\ttotal: 561ms\tremaining: 1.04s\n",
      "350:\tlearn: 0.3270682\ttotal: 562ms\tremaining: 1.04s\n",
      "351:\tlearn: 0.3268660\ttotal: 563ms\tremaining: 1.04s\n",
      "352:\tlearn: 0.3266701\ttotal: 564ms\tremaining: 1.03s\n",
      "353:\tlearn: 0.3265748\ttotal: 565ms\tremaining: 1.03s\n",
      "354:\tlearn: 0.3263575\ttotal: 566ms\tremaining: 1.03s\n",
      "355:\tlearn: 0.3261232\ttotal: 567ms\tremaining: 1.03s\n",
      "356:\tlearn: 0.3258959\ttotal: 568ms\tremaining: 1.02s\n",
      "357:\tlearn: 0.3257788\ttotal: 569ms\tremaining: 1.02s\n",
      "358:\tlearn: 0.3255893\ttotal: 570ms\tremaining: 1.02s\n",
      "359:\tlearn: 0.3253664\ttotal: 571ms\tremaining: 1.01s\n",
      "360:\tlearn: 0.3252371\ttotal: 572ms\tremaining: 1.01s\n",
      "361:\tlearn: 0.3250468\ttotal: 573ms\tremaining: 1.01s\n",
      "362:\tlearn: 0.3248758\ttotal: 574ms\tremaining: 1.01s\n",
      "363:\tlearn: 0.3247652\ttotal: 575ms\tremaining: 1s\n",
      "364:\tlearn: 0.3247403\ttotal: 576ms\tremaining: 1s\n",
      "365:\tlearn: 0.3243928\ttotal: 577ms\tremaining: 1000ms\n",
      "366:\tlearn: 0.3243343\ttotal: 578ms\tremaining: 997ms\n",
      "367:\tlearn: 0.3243221\ttotal: 579ms\tremaining: 994ms\n",
      "368:\tlearn: 0.3241893\ttotal: 580ms\tremaining: 991ms\n",
      "369:\tlearn: 0.3241577\ttotal: 580ms\tremaining: 988ms\n",
      "370:\tlearn: 0.3240481\ttotal: 582ms\tremaining: 986ms\n",
      "371:\tlearn: 0.3239396\ttotal: 583ms\tremaining: 984ms\n",
      "372:\tlearn: 0.3237781\ttotal: 584ms\tremaining: 982ms\n",
      "373:\tlearn: 0.3236733\ttotal: 585ms\tremaining: 979ms\n",
      "374:\tlearn: 0.3235889\ttotal: 586ms\tremaining: 977ms\n",
      "375:\tlearn: 0.3233586\ttotal: 587ms\tremaining: 974ms\n",
      "376:\tlearn: 0.3231252\ttotal: 588ms\tremaining: 972ms\n",
      "377:\tlearn: 0.3229567\ttotal: 589ms\tremaining: 970ms\n",
      "378:\tlearn: 0.3228761\ttotal: 590ms\tremaining: 967ms\n",
      "379:\tlearn: 0.3226378\ttotal: 591ms\tremaining: 965ms\n",
      "380:\tlearn: 0.3226124\ttotal: 592ms\tremaining: 962ms\n",
      "381:\tlearn: 0.3224442\ttotal: 593ms\tremaining: 960ms\n",
      "382:\tlearn: 0.3223747\ttotal: 594ms\tremaining: 957ms\n",
      "383:\tlearn: 0.3222752\ttotal: 595ms\tremaining: 955ms\n",
      "384:\tlearn: 0.3221238\ttotal: 596ms\tremaining: 952ms\n",
      "385:\tlearn: 0.3219097\ttotal: 597ms\tremaining: 950ms\n",
      "386:\tlearn: 0.3218015\ttotal: 599ms\tremaining: 948ms\n",
      "387:\tlearn: 0.3215443\ttotal: 600ms\tremaining: 946ms\n",
      "388:\tlearn: 0.3213962\ttotal: 601ms\tremaining: 944ms\n",
      "389:\tlearn: 0.3213668\ttotal: 602ms\tremaining: 942ms\n",
      "390:\tlearn: 0.3212285\ttotal: 603ms\tremaining: 939ms\n",
      "391:\tlearn: 0.3210017\ttotal: 604ms\tremaining: 937ms\n",
      "392:\tlearn: 0.3207542\ttotal: 605ms\tremaining: 934ms\n",
      "393:\tlearn: 0.3204902\ttotal: 606ms\tremaining: 932ms\n",
      "394:\tlearn: 0.3202627\ttotal: 607ms\tremaining: 930ms\n",
      "395:\tlearn: 0.3200657\ttotal: 608ms\tremaining: 928ms\n",
      "396:\tlearn: 0.3197727\ttotal: 609ms\tremaining: 925ms\n",
      "397:\tlearn: 0.3196934\ttotal: 610ms\tremaining: 923ms\n",
      "398:\tlearn: 0.3195296\ttotal: 611ms\tremaining: 920ms\n",
      "399:\tlearn: 0.3194697\ttotal: 612ms\tremaining: 918ms\n",
      "400:\tlearn: 0.3193837\ttotal: 613ms\tremaining: 916ms\n",
      "401:\tlearn: 0.3193336\ttotal: 614ms\tremaining: 913ms\n",
      "402:\tlearn: 0.3192256\ttotal: 615ms\tremaining: 910ms\n",
      "403:\tlearn: 0.3190788\ttotal: 616ms\tremaining: 908ms\n",
      "404:\tlearn: 0.3189177\ttotal: 617ms\tremaining: 906ms\n",
      "405:\tlearn: 0.3187390\ttotal: 618ms\tremaining: 904ms\n",
      "406:\tlearn: 0.3185300\ttotal: 619ms\tremaining: 901ms\n",
      "407:\tlearn: 0.3183823\ttotal: 620ms\tremaining: 899ms\n",
      "408:\tlearn: 0.3181672\ttotal: 621ms\tremaining: 897ms\n",
      "409:\tlearn: 0.3178911\ttotal: 622ms\tremaining: 895ms\n",
      "410:\tlearn: 0.3177896\ttotal: 623ms\tremaining: 892ms\n",
      "411:\tlearn: 0.3177635\ttotal: 623ms\tremaining: 890ms\n",
      "412:\tlearn: 0.3175092\ttotal: 624ms\tremaining: 887ms\n",
      "413:\tlearn: 0.3173806\ttotal: 625ms\tremaining: 885ms\n",
      "414:\tlearn: 0.3171642\ttotal: 626ms\tremaining: 883ms\n",
      "415:\tlearn: 0.3168919\ttotal: 627ms\tremaining: 881ms\n",
      "416:\tlearn: 0.3166320\ttotal: 628ms\tremaining: 879ms\n",
      "417:\tlearn: 0.3164688\ttotal: 629ms\tremaining: 876ms\n",
      "418:\tlearn: 0.3163313\ttotal: 630ms\tremaining: 874ms\n",
      "419:\tlearn: 0.3162629\ttotal: 631ms\tremaining: 872ms\n",
      "420:\tlearn: 0.3161897\ttotal: 632ms\tremaining: 870ms\n",
      "421:\tlearn: 0.3158241\ttotal: 633ms\tremaining: 868ms\n",
      "422:\tlearn: 0.3157041\ttotal: 635ms\tremaining: 866ms\n",
      "423:\tlearn: 0.3154946\ttotal: 636ms\tremaining: 864ms\n",
      "424:\tlearn: 0.3152653\ttotal: 637ms\tremaining: 861ms\n",
      "425:\tlearn: 0.3151932\ttotal: 638ms\tremaining: 859ms\n",
      "426:\tlearn: 0.3149621\ttotal: 639ms\tremaining: 857ms\n",
      "427:\tlearn: 0.3148137\ttotal: 640ms\tremaining: 855ms\n",
      "428:\tlearn: 0.3146853\ttotal: 641ms\tremaining: 853ms\n",
      "429:\tlearn: 0.3145723\ttotal: 642ms\tremaining: 851ms\n",
      "430:\tlearn: 0.3144883\ttotal: 644ms\tremaining: 850ms\n",
      "431:\tlearn: 0.3142902\ttotal: 645ms\tremaining: 848ms\n",
      "432:\tlearn: 0.3142303\ttotal: 646ms\tremaining: 846ms\n",
      "433:\tlearn: 0.3141377\ttotal: 647ms\tremaining: 843ms\n",
      "434:\tlearn: 0.3140161\ttotal: 648ms\tremaining: 841ms\n",
      "435:\tlearn: 0.3138909\ttotal: 649ms\tremaining: 839ms\n",
      "436:\tlearn: 0.3136637\ttotal: 650ms\tremaining: 838ms\n",
      "437:\tlearn: 0.3135557\ttotal: 651ms\tremaining: 836ms\n",
      "438:\tlearn: 0.3134545\ttotal: 652ms\tremaining: 833ms\n",
      "439:\tlearn: 0.3134393\ttotal: 653ms\tremaining: 831ms\n",
      "440:\tlearn: 0.3133888\ttotal: 654ms\tremaining: 829ms\n",
      "441:\tlearn: 0.3132995\ttotal: 655ms\tremaining: 827ms\n",
      "442:\tlearn: 0.3131442\ttotal: 656ms\tremaining: 825ms\n",
      "443:\tlearn: 0.3130382\ttotal: 657ms\tremaining: 823ms\n",
      "444:\tlearn: 0.3129489\ttotal: 658ms\tremaining: 821ms\n",
      "445:\tlearn: 0.3128960\ttotal: 659ms\tremaining: 819ms\n",
      "446:\tlearn: 0.3127145\ttotal: 661ms\tremaining: 818ms\n",
      "447:\tlearn: 0.3125959\ttotal: 662ms\tremaining: 816ms\n",
      "448:\tlearn: 0.3125919\ttotal: 663ms\tremaining: 814ms\n",
      "449:\tlearn: 0.3124815\ttotal: 664ms\tremaining: 812ms\n",
      "450:\tlearn: 0.3123323\ttotal: 665ms\tremaining: 810ms\n",
      "451:\tlearn: 0.3122396\ttotal: 666ms\tremaining: 807ms\n",
      "452:\tlearn: 0.3120755\ttotal: 667ms\tremaining: 805ms\n",
      "453:\tlearn: 0.3119721\ttotal: 668ms\tremaining: 803ms\n",
      "454:\tlearn: 0.3118337\ttotal: 669ms\tremaining: 801ms\n",
      "455:\tlearn: 0.3117352\ttotal: 670ms\tremaining: 799ms\n",
      "456:\tlearn: 0.3116560\ttotal: 671ms\tremaining: 797ms\n",
      "457:\tlearn: 0.3114432\ttotal: 672ms\tremaining: 795ms\n",
      "458:\tlearn: 0.3112807\ttotal: 673ms\tremaining: 793ms\n",
      "459:\tlearn: 0.3110543\ttotal: 674ms\tremaining: 791ms\n",
      "460:\tlearn: 0.3109123\ttotal: 675ms\tremaining: 789ms\n",
      "461:\tlearn: 0.3107861\ttotal: 676ms\tremaining: 787ms\n",
      "462:\tlearn: 0.3106145\ttotal: 677ms\tremaining: 785ms\n",
      "463:\tlearn: 0.3105384\ttotal: 678ms\tremaining: 783ms\n",
      "464:\tlearn: 0.3104834\ttotal: 679ms\tremaining: 781ms\n",
      "465:\tlearn: 0.3104144\ttotal: 680ms\tremaining: 779ms\n",
      "466:\tlearn: 0.3102801\ttotal: 681ms\tremaining: 777ms\n",
      "467:\tlearn: 0.3101780\ttotal: 682ms\tremaining: 775ms\n",
      "468:\tlearn: 0.3100771\ttotal: 683ms\tremaining: 773ms\n",
      "469:\tlearn: 0.3098800\ttotal: 684ms\tremaining: 771ms\n",
      "470:\tlearn: 0.3098158\ttotal: 685ms\tremaining: 769ms\n",
      "471:\tlearn: 0.3097560\ttotal: 686ms\tremaining: 767ms\n",
      "472:\tlearn: 0.3095755\ttotal: 687ms\tremaining: 765ms\n",
      "473:\tlearn: 0.3095327\ttotal: 688ms\tremaining: 763ms\n",
      "474:\tlearn: 0.3093303\ttotal: 689ms\tremaining: 762ms\n",
      "475:\tlearn: 0.3092150\ttotal: 690ms\tremaining: 760ms\n",
      "476:\tlearn: 0.3090890\ttotal: 691ms\tremaining: 758ms\n",
      "477:\tlearn: 0.3090125\ttotal: 692ms\tremaining: 756ms\n",
      "478:\tlearn: 0.3089181\ttotal: 693ms\tremaining: 754ms\n",
      "479:\tlearn: 0.3088336\ttotal: 694ms\tremaining: 752ms\n",
      "480:\tlearn: 0.3085799\ttotal: 695ms\tremaining: 750ms\n",
      "481:\tlearn: 0.3083691\ttotal: 696ms\tremaining: 748ms\n",
      "482:\tlearn: 0.3082406\ttotal: 697ms\tremaining: 746ms\n",
      "483:\tlearn: 0.3082215\ttotal: 698ms\tremaining: 745ms\n",
      "484:\tlearn: 0.3081138\ttotal: 699ms\tremaining: 743ms\n",
      "485:\tlearn: 0.3079897\ttotal: 700ms\tremaining: 741ms\n",
      "486:\tlearn: 0.3078718\ttotal: 701ms\tremaining: 739ms\n",
      "487:\tlearn: 0.3077215\ttotal: 702ms\tremaining: 737ms\n",
      "488:\tlearn: 0.3077010\ttotal: 703ms\tremaining: 735ms\n",
      "489:\tlearn: 0.3074744\ttotal: 704ms\tremaining: 733ms\n",
      "490:\tlearn: 0.3072634\ttotal: 705ms\tremaining: 731ms\n",
      "491:\tlearn: 0.3072475\ttotal: 706ms\tremaining: 728ms\n",
      "492:\tlearn: 0.3071595\ttotal: 707ms\tremaining: 727ms\n",
      "493:\tlearn: 0.3070410\ttotal: 708ms\tremaining: 725ms\n",
      "494:\tlearn: 0.3068810\ttotal: 709ms\tremaining: 723ms\n",
      "495:\tlearn: 0.3067315\ttotal: 710ms\tremaining: 721ms\n",
      "496:\tlearn: 0.3064770\ttotal: 711ms\tremaining: 719ms\n",
      "497:\tlearn: 0.3064467\ttotal: 712ms\tremaining: 717ms\n",
      "498:\tlearn: 0.3063002\ttotal: 713ms\tremaining: 715ms\n",
      "499:\tlearn: 0.3062847\ttotal: 713ms\tremaining: 713ms\n",
      "500:\tlearn: 0.3062051\ttotal: 714ms\tremaining: 711ms\n",
      "501:\tlearn: 0.3060788\ttotal: 715ms\tremaining: 709ms\n",
      "502:\tlearn: 0.3059827\ttotal: 716ms\tremaining: 708ms\n",
      "503:\tlearn: 0.3058498\ttotal: 717ms\tremaining: 706ms\n",
      "504:\tlearn: 0.3058043\ttotal: 718ms\tremaining: 704ms\n",
      "505:\tlearn: 0.3056431\ttotal: 720ms\tremaining: 703ms\n",
      "506:\tlearn: 0.3055885\ttotal: 721ms\tremaining: 701ms\n",
      "507:\tlearn: 0.3055687\ttotal: 722ms\tremaining: 699ms\n",
      "508:\tlearn: 0.3053257\ttotal: 723ms\tremaining: 697ms\n",
      "509:\tlearn: 0.3051183\ttotal: 724ms\tremaining: 695ms\n",
      "510:\tlearn: 0.3050174\ttotal: 725ms\tremaining: 693ms\n",
      "511:\tlearn: 0.3049028\ttotal: 726ms\tremaining: 692ms\n",
      "512:\tlearn: 0.3047795\ttotal: 727ms\tremaining: 690ms\n",
      "513:\tlearn: 0.3046918\ttotal: 728ms\tremaining: 688ms\n",
      "514:\tlearn: 0.3045672\ttotal: 729ms\tremaining: 686ms\n",
      "515:\tlearn: 0.3044681\ttotal: 730ms\tremaining: 684ms\n",
      "516:\tlearn: 0.3044039\ttotal: 731ms\tremaining: 683ms\n",
      "517:\tlearn: 0.3043545\ttotal: 732ms\tremaining: 681ms\n",
      "518:\tlearn: 0.3042007\ttotal: 733ms\tremaining: 679ms\n",
      "519:\tlearn: 0.3041054\ttotal: 734ms\tremaining: 677ms\n",
      "520:\tlearn: 0.3040072\ttotal: 735ms\tremaining: 676ms\n",
      "521:\tlearn: 0.3039845\ttotal: 735ms\tremaining: 673ms\n",
      "522:\tlearn: 0.3039044\ttotal: 736ms\tremaining: 671ms\n",
      "523:\tlearn: 0.3037900\ttotal: 737ms\tremaining: 670ms\n",
      "524:\tlearn: 0.3037856\ttotal: 738ms\tremaining: 667ms\n",
      "525:\tlearn: 0.3037126\ttotal: 739ms\tremaining: 666ms\n",
      "526:\tlearn: 0.3035184\ttotal: 740ms\tremaining: 664ms\n",
      "527:\tlearn: 0.3034185\ttotal: 741ms\tremaining: 662ms\n",
      "528:\tlearn: 0.3032301\ttotal: 742ms\tremaining: 661ms\n",
      "529:\tlearn: 0.3031039\ttotal: 743ms\tremaining: 659ms\n",
      "530:\tlearn: 0.3030460\ttotal: 744ms\tremaining: 657ms\n",
      "531:\tlearn: 0.3029432\ttotal: 745ms\tremaining: 656ms\n",
      "532:\tlearn: 0.3027806\ttotal: 747ms\tremaining: 654ms\n",
      "533:\tlearn: 0.3027357\ttotal: 748ms\tremaining: 652ms\n",
      "534:\tlearn: 0.3026515\ttotal: 749ms\tremaining: 651ms\n",
      "535:\tlearn: 0.3025344\ttotal: 750ms\tremaining: 649ms\n",
      "536:\tlearn: 0.3024547\ttotal: 751ms\tremaining: 647ms\n",
      "537:\tlearn: 0.3023828\ttotal: 752ms\tremaining: 645ms\n",
      "538:\tlearn: 0.3022340\ttotal: 753ms\tremaining: 644ms\n",
      "539:\tlearn: 0.3020917\ttotal: 754ms\tremaining: 642ms\n",
      "540:\tlearn: 0.3019061\ttotal: 755ms\tremaining: 640ms\n",
      "541:\tlearn: 0.3018727\ttotal: 756ms\tremaining: 638ms\n",
      "542:\tlearn: 0.3018060\ttotal: 757ms\tremaining: 637ms\n",
      "543:\tlearn: 0.3017484\ttotal: 758ms\tremaining: 635ms\n",
      "544:\tlearn: 0.3016423\ttotal: 759ms\tremaining: 633ms\n",
      "545:\tlearn: 0.3014926\ttotal: 760ms\tremaining: 632ms\n",
      "546:\tlearn: 0.3014121\ttotal: 760ms\tremaining: 630ms\n",
      "547:\tlearn: 0.3012626\ttotal: 761ms\tremaining: 628ms\n",
      "548:\tlearn: 0.3011252\ttotal: 762ms\tremaining: 626ms\n",
      "549:\tlearn: 0.3010267\ttotal: 763ms\tremaining: 625ms\n",
      "550:\tlearn: 0.3008311\ttotal: 764ms\tremaining: 623ms\n",
      "551:\tlearn: 0.3006674\ttotal: 765ms\tremaining: 621ms\n",
      "552:\tlearn: 0.3005445\ttotal: 766ms\tremaining: 619ms\n",
      "553:\tlearn: 0.3004625\ttotal: 767ms\tremaining: 618ms\n",
      "554:\tlearn: 0.3002814\ttotal: 768ms\tremaining: 616ms\n",
      "555:\tlearn: 0.3002137\ttotal: 769ms\tremaining: 614ms\n",
      "556:\tlearn: 0.3000040\ttotal: 770ms\tremaining: 613ms\n",
      "557:\tlearn: 0.2997805\ttotal: 771ms\tremaining: 611ms\n",
      "558:\tlearn: 0.2996651\ttotal: 772ms\tremaining: 609ms\n",
      "559:\tlearn: 0.2995484\ttotal: 773ms\tremaining: 608ms\n",
      "560:\tlearn: 0.2994378\ttotal: 775ms\tremaining: 606ms\n",
      "561:\tlearn: 0.2993556\ttotal: 776ms\tremaining: 605ms\n",
      "562:\tlearn: 0.2993205\ttotal: 777ms\tremaining: 603ms\n",
      "563:\tlearn: 0.2992062\ttotal: 778ms\tremaining: 601ms\n",
      "564:\tlearn: 0.2991306\ttotal: 779ms\tremaining: 600ms\n",
      "565:\tlearn: 0.2989895\ttotal: 780ms\tremaining: 598ms\n",
      "566:\tlearn: 0.2988683\ttotal: 781ms\tremaining: 597ms\n",
      "567:\tlearn: 0.2987492\ttotal: 782ms\tremaining: 595ms\n",
      "568:\tlearn: 0.2987148\ttotal: 783ms\tremaining: 593ms\n",
      "569:\tlearn: 0.2986751\ttotal: 784ms\tremaining: 592ms\n",
      "570:\tlearn: 0.2985473\ttotal: 785ms\tremaining: 590ms\n",
      "571:\tlearn: 0.2985378\ttotal: 786ms\tremaining: 588ms\n",
      "572:\tlearn: 0.2983068\ttotal: 787ms\tremaining: 586ms\n",
      "573:\tlearn: 0.2981665\ttotal: 788ms\tremaining: 585ms\n",
      "574:\tlearn: 0.2980186\ttotal: 789ms\tremaining: 583ms\n",
      "575:\tlearn: 0.2979872\ttotal: 790ms\tremaining: 582ms\n",
      "576:\tlearn: 0.2978643\ttotal: 791ms\tremaining: 580ms\n",
      "577:\tlearn: 0.2977196\ttotal: 792ms\tremaining: 578ms\n",
      "578:\tlearn: 0.2976289\ttotal: 793ms\tremaining: 577ms\n",
      "579:\tlearn: 0.2975509\ttotal: 795ms\tremaining: 575ms\n",
      "580:\tlearn: 0.2973543\ttotal: 796ms\tremaining: 574ms\n",
      "581:\tlearn: 0.2971403\ttotal: 797ms\tremaining: 572ms\n",
      "582:\tlearn: 0.2969255\ttotal: 798ms\tremaining: 571ms\n",
      "583:\tlearn: 0.2968716\ttotal: 799ms\tremaining: 570ms\n",
      "584:\tlearn: 0.2967577\ttotal: 801ms\tremaining: 568ms\n",
      "585:\tlearn: 0.2965096\ttotal: 803ms\tremaining: 567ms\n",
      "586:\tlearn: 0.2964660\ttotal: 804ms\tremaining: 566ms\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "587:\tlearn: 0.2964541\ttotal: 805ms\tremaining: 564ms\n",
      "588:\tlearn: 0.2963067\ttotal: 807ms\tremaining: 563ms\n",
      "589:\tlearn: 0.2961962\ttotal: 808ms\tremaining: 561ms\n",
      "590:\tlearn: 0.2961603\ttotal: 809ms\tremaining: 560ms\n",
      "591:\tlearn: 0.2961469\ttotal: 809ms\tremaining: 558ms\n",
      "592:\tlearn: 0.2960685\ttotal: 810ms\tremaining: 556ms\n",
      "593:\tlearn: 0.2958687\ttotal: 812ms\tremaining: 555ms\n",
      "594:\tlearn: 0.2957421\ttotal: 813ms\tremaining: 553ms\n",
      "595:\tlearn: 0.2956312\ttotal: 814ms\tremaining: 552ms\n",
      "596:\tlearn: 0.2955434\ttotal: 816ms\tremaining: 551ms\n",
      "597:\tlearn: 0.2952732\ttotal: 817ms\tremaining: 549ms\n",
      "598:\tlearn: 0.2952468\ttotal: 818ms\tremaining: 547ms\n",
      "599:\tlearn: 0.2950704\ttotal: 819ms\tremaining: 546ms\n",
      "600:\tlearn: 0.2948845\ttotal: 820ms\tremaining: 544ms\n",
      "601:\tlearn: 0.2947693\ttotal: 821ms\tremaining: 543ms\n",
      "602:\tlearn: 0.2945863\ttotal: 822ms\tremaining: 541ms\n",
      "603:\tlearn: 0.2944739\ttotal: 824ms\tremaining: 540ms\n",
      "604:\tlearn: 0.2943475\ttotal: 825ms\tremaining: 539ms\n",
      "605:\tlearn: 0.2942942\ttotal: 826ms\tremaining: 537ms\n",
      "606:\tlearn: 0.2942753\ttotal: 827ms\tremaining: 536ms\n",
      "607:\tlearn: 0.2941945\ttotal: 828ms\tremaining: 534ms\n",
      "608:\tlearn: 0.2940892\ttotal: 830ms\tremaining: 533ms\n",
      "609:\tlearn: 0.2940454\ttotal: 831ms\tremaining: 531ms\n",
      "610:\tlearn: 0.2940342\ttotal: 831ms\tremaining: 529ms\n",
      "611:\tlearn: 0.2938592\ttotal: 832ms\tremaining: 528ms\n",
      "612:\tlearn: 0.2937729\ttotal: 834ms\tremaining: 526ms\n",
      "613:\tlearn: 0.2936799\ttotal: 835ms\tremaining: 525ms\n",
      "614:\tlearn: 0.2935277\ttotal: 836ms\tremaining: 523ms\n",
      "615:\tlearn: 0.2934158\ttotal: 837ms\tremaining: 522ms\n",
      "616:\tlearn: 0.2932758\ttotal: 838ms\tremaining: 520ms\n",
      "617:\tlearn: 0.2931298\ttotal: 840ms\tremaining: 519ms\n",
      "618:\tlearn: 0.2929937\ttotal: 841ms\tremaining: 517ms\n",
      "619:\tlearn: 0.2929035\ttotal: 842ms\tremaining: 516ms\n",
      "620:\tlearn: 0.2928774\ttotal: 843ms\tremaining: 514ms\n",
      "621:\tlearn: 0.2926599\ttotal: 844ms\tremaining: 513ms\n",
      "622:\tlearn: 0.2923901\ttotal: 845ms\tremaining: 511ms\n",
      "623:\tlearn: 0.2923545\ttotal: 846ms\tremaining: 510ms\n",
      "624:\tlearn: 0.2923015\ttotal: 847ms\tremaining: 508ms\n",
      "625:\tlearn: 0.2922130\ttotal: 848ms\tremaining: 507ms\n",
      "626:\tlearn: 0.2920696\ttotal: 849ms\tremaining: 505ms\n",
      "627:\tlearn: 0.2919205\ttotal: 850ms\tremaining: 504ms\n",
      "628:\tlearn: 0.2918231\ttotal: 851ms\tremaining: 502ms\n",
      "629:\tlearn: 0.2916690\ttotal: 853ms\tremaining: 501ms\n",
      "630:\tlearn: 0.2916174\ttotal: 854ms\tremaining: 499ms\n",
      "631:\tlearn: 0.2915610\ttotal: 855ms\tremaining: 498ms\n",
      "632:\tlearn: 0.2914977\ttotal: 856ms\tremaining: 496ms\n",
      "633:\tlearn: 0.2913428\ttotal: 857ms\tremaining: 495ms\n",
      "634:\tlearn: 0.2911908\ttotal: 858ms\tremaining: 493ms\n",
      "635:\tlearn: 0.2910728\ttotal: 859ms\tremaining: 492ms\n",
      "636:\tlearn: 0.2909249\ttotal: 860ms\tremaining: 490ms\n",
      "637:\tlearn: 0.2908024\ttotal: 862ms\tremaining: 489ms\n",
      "638:\tlearn: 0.2906880\ttotal: 863ms\tremaining: 487ms\n",
      "639:\tlearn: 0.2905559\ttotal: 864ms\tremaining: 486ms\n",
      "640:\tlearn: 0.2905140\ttotal: 865ms\tremaining: 484ms\n",
      "641:\tlearn: 0.2902851\ttotal: 866ms\tremaining: 483ms\n",
      "642:\tlearn: 0.2901904\ttotal: 867ms\tremaining: 481ms\n",
      "643:\tlearn: 0.2900720\ttotal: 868ms\tremaining: 480ms\n",
      "644:\tlearn: 0.2899198\ttotal: 869ms\tremaining: 479ms\n",
      "645:\tlearn: 0.2898092\ttotal: 870ms\tremaining: 477ms\n",
      "646:\tlearn: 0.2896641\ttotal: 872ms\tremaining: 476ms\n",
      "647:\tlearn: 0.2895770\ttotal: 873ms\tremaining: 474ms\n",
      "648:\tlearn: 0.2894452\ttotal: 874ms\tremaining: 473ms\n",
      "649:\tlearn: 0.2893049\ttotal: 876ms\tremaining: 472ms\n",
      "650:\tlearn: 0.2892829\ttotal: 878ms\tremaining: 471ms\n",
      "651:\tlearn: 0.2891726\ttotal: 879ms\tremaining: 469ms\n",
      "652:\tlearn: 0.2890985\ttotal: 881ms\tremaining: 468ms\n",
      "653:\tlearn: 0.2890446\ttotal: 882ms\tremaining: 466ms\n",
      "654:\tlearn: 0.2889486\ttotal: 883ms\tremaining: 465ms\n",
      "655:\tlearn: 0.2888768\ttotal: 884ms\tremaining: 463ms\n",
      "656:\tlearn: 0.2887969\ttotal: 885ms\tremaining: 462ms\n",
      "657:\tlearn: 0.2886840\ttotal: 886ms\tremaining: 460ms\n",
      "658:\tlearn: 0.2886132\ttotal: 887ms\tremaining: 459ms\n",
      "659:\tlearn: 0.2885347\ttotal: 888ms\tremaining: 457ms\n",
      "660:\tlearn: 0.2884952\ttotal: 889ms\tremaining: 456ms\n",
      "661:\tlearn: 0.2883412\ttotal: 890ms\tremaining: 454ms\n",
      "662:\tlearn: 0.2882455\ttotal: 891ms\tremaining: 453ms\n",
      "663:\tlearn: 0.2881157\ttotal: 892ms\tremaining: 452ms\n",
      "664:\tlearn: 0.2879658\ttotal: 893ms\tremaining: 450ms\n",
      "665:\tlearn: 0.2878309\ttotal: 895ms\tremaining: 449ms\n",
      "666:\tlearn: 0.2877590\ttotal: 896ms\tremaining: 447ms\n",
      "667:\tlearn: 0.2875809\ttotal: 897ms\tremaining: 446ms\n",
      "668:\tlearn: 0.2874602\ttotal: 898ms\tremaining: 444ms\n",
      "669:\tlearn: 0.2873926\ttotal: 899ms\tremaining: 443ms\n",
      "670:\tlearn: 0.2873015\ttotal: 900ms\tremaining: 441ms\n",
      "671:\tlearn: 0.2872138\ttotal: 901ms\tremaining: 440ms\n",
      "672:\tlearn: 0.2871335\ttotal: 902ms\tremaining: 438ms\n",
      "673:\tlearn: 0.2869738\ttotal: 903ms\tremaining: 437ms\n",
      "674:\tlearn: 0.2867912\ttotal: 905ms\tremaining: 436ms\n",
      "675:\tlearn: 0.2867361\ttotal: 906ms\tremaining: 434ms\n",
      "676:\tlearn: 0.2864418\ttotal: 907ms\tremaining: 433ms\n",
      "677:\tlearn: 0.2862719\ttotal: 908ms\tremaining: 431ms\n",
      "678:\tlearn: 0.2861712\ttotal: 909ms\tremaining: 430ms\n",
      "679:\tlearn: 0.2860157\ttotal: 910ms\tremaining: 428ms\n",
      "680:\tlearn: 0.2857729\ttotal: 912ms\tremaining: 427ms\n",
      "681:\tlearn: 0.2856815\ttotal: 913ms\tremaining: 426ms\n",
      "682:\tlearn: 0.2854762\ttotal: 914ms\tremaining: 424ms\n",
      "683:\tlearn: 0.2854134\ttotal: 915ms\tremaining: 423ms\n",
      "684:\tlearn: 0.2853347\ttotal: 916ms\tremaining: 421ms\n",
      "685:\tlearn: 0.2852855\ttotal: 917ms\tremaining: 420ms\n",
      "686:\tlearn: 0.2850584\ttotal: 918ms\tremaining: 418ms\n",
      "687:\tlearn: 0.2847548\ttotal: 919ms\tremaining: 417ms\n",
      "688:\tlearn: 0.2846957\ttotal: 920ms\tremaining: 415ms\n",
      "689:\tlearn: 0.2845469\ttotal: 922ms\tremaining: 414ms\n",
      "690:\tlearn: 0.2844987\ttotal: 923ms\tremaining: 413ms\n",
      "691:\tlearn: 0.2843161\ttotal: 924ms\tremaining: 411ms\n",
      "692:\tlearn: 0.2842032\ttotal: 925ms\tremaining: 410ms\n",
      "693:\tlearn: 0.2841340\ttotal: 926ms\tremaining: 408ms\n",
      "694:\tlearn: 0.2840410\ttotal: 928ms\tremaining: 407ms\n",
      "695:\tlearn: 0.2838450\ttotal: 929ms\tremaining: 406ms\n",
      "696:\tlearn: 0.2837900\ttotal: 930ms\tremaining: 404ms\n",
      "697:\tlearn: 0.2837462\ttotal: 931ms\tremaining: 403ms\n",
      "698:\tlearn: 0.2835805\ttotal: 932ms\tremaining: 401ms\n",
      "699:\tlearn: 0.2835224\ttotal: 933ms\tremaining: 400ms\n",
      "700:\tlearn: 0.2833545\ttotal: 935ms\tremaining: 399ms\n",
      "701:\tlearn: 0.2832162\ttotal: 936ms\tremaining: 397ms\n",
      "702:\tlearn: 0.2831633\ttotal: 938ms\tremaining: 396ms\n",
      "703:\tlearn: 0.2831206\ttotal: 939ms\tremaining: 395ms\n",
      "704:\tlearn: 0.2829767\ttotal: 940ms\tremaining: 393ms\n",
      "705:\tlearn: 0.2829667\ttotal: 941ms\tremaining: 392ms\n",
      "706:\tlearn: 0.2827761\ttotal: 942ms\tremaining: 390ms\n",
      "707:\tlearn: 0.2824377\ttotal: 943ms\tremaining: 389ms\n",
      "708:\tlearn: 0.2823739\ttotal: 944ms\tremaining: 387ms\n",
      "709:\tlearn: 0.2821424\ttotal: 945ms\tremaining: 386ms\n",
      "710:\tlearn: 0.2820554\ttotal: 946ms\tremaining: 385ms\n",
      "711:\tlearn: 0.2819645\ttotal: 947ms\tremaining: 383ms\n",
      "712:\tlearn: 0.2818310\ttotal: 949ms\tremaining: 382ms\n",
      "713:\tlearn: 0.2817026\ttotal: 950ms\tremaining: 380ms\n",
      "714:\tlearn: 0.2816223\ttotal: 951ms\tremaining: 379ms\n",
      "715:\tlearn: 0.2815874\ttotal: 953ms\tremaining: 378ms\n",
      "716:\tlearn: 0.2815653\ttotal: 953ms\tremaining: 376ms\n",
      "717:\tlearn: 0.2815240\ttotal: 954ms\tremaining: 375ms\n",
      "718:\tlearn: 0.2814851\ttotal: 955ms\tremaining: 373ms\n",
      "719:\tlearn: 0.2814000\ttotal: 957ms\tremaining: 372ms\n",
      "720:\tlearn: 0.2812436\ttotal: 958ms\tremaining: 371ms\n",
      "721:\tlearn: 0.2811408\ttotal: 959ms\tremaining: 369ms\n",
      "722:\tlearn: 0.2810318\ttotal: 961ms\tremaining: 368ms\n",
      "723:\tlearn: 0.2809471\ttotal: 962ms\tremaining: 367ms\n",
      "724:\tlearn: 0.2808964\ttotal: 963ms\tremaining: 365ms\n",
      "725:\tlearn: 0.2807185\ttotal: 964ms\tremaining: 364ms\n",
      "726:\tlearn: 0.2806533\ttotal: 965ms\tremaining: 362ms\n",
      "727:\tlearn: 0.2805334\ttotal: 966ms\tremaining: 361ms\n",
      "728:\tlearn: 0.2804354\ttotal: 967ms\tremaining: 359ms\n",
      "729:\tlearn: 0.2803634\ttotal: 968ms\tremaining: 358ms\n",
      "730:\tlearn: 0.2802542\ttotal: 970ms\tremaining: 357ms\n",
      "731:\tlearn: 0.2801831\ttotal: 971ms\tremaining: 355ms\n",
      "732:\tlearn: 0.2799985\ttotal: 972ms\tremaining: 354ms\n",
      "733:\tlearn: 0.2798394\ttotal: 973ms\tremaining: 353ms\n",
      "734:\tlearn: 0.2798038\ttotal: 974ms\tremaining: 351ms\n",
      "735:\tlearn: 0.2796865\ttotal: 976ms\tremaining: 350ms\n",
      "736:\tlearn: 0.2796094\ttotal: 977ms\tremaining: 349ms\n",
      "737:\tlearn: 0.2795135\ttotal: 978ms\tremaining: 347ms\n",
      "738:\tlearn: 0.2794188\ttotal: 979ms\tremaining: 346ms\n",
      "739:\tlearn: 0.2792592\ttotal: 980ms\tremaining: 344ms\n",
      "740:\tlearn: 0.2791876\ttotal: 981ms\tremaining: 343ms\n",
      "741:\tlearn: 0.2791323\ttotal: 982ms\tremaining: 341ms\n",
      "742:\tlearn: 0.2789331\ttotal: 983ms\tremaining: 340ms\n",
      "743:\tlearn: 0.2786663\ttotal: 984ms\tremaining: 339ms\n",
      "744:\tlearn: 0.2785949\ttotal: 986ms\tremaining: 337ms\n",
      "745:\tlearn: 0.2785503\ttotal: 987ms\tremaining: 336ms\n",
      "746:\tlearn: 0.2783982\ttotal: 988ms\tremaining: 335ms\n",
      "747:\tlearn: 0.2782860\ttotal: 990ms\tremaining: 333ms\n",
      "748:\tlearn: 0.2781833\ttotal: 991ms\tremaining: 332ms\n",
      "749:\tlearn: 0.2781145\ttotal: 992ms\tremaining: 331ms\n",
      "750:\tlearn: 0.2779833\ttotal: 993ms\tremaining: 329ms\n",
      "751:\tlearn: 0.2779069\ttotal: 994ms\tremaining: 328ms\n",
      "752:\tlearn: 0.2777634\ttotal: 995ms\tremaining: 326ms\n",
      "753:\tlearn: 0.2776630\ttotal: 996ms\tremaining: 325ms\n",
      "754:\tlearn: 0.2775628\ttotal: 997ms\tremaining: 324ms\n",
      "755:\tlearn: 0.2774674\ttotal: 998ms\tremaining: 322ms\n",
      "756:\tlearn: 0.2773710\ttotal: 1s\tremaining: 321ms\n",
      "757:\tlearn: 0.2771739\ttotal: 1s\tremaining: 320ms\n",
      "758:\tlearn: 0.2770339\ttotal: 1s\tremaining: 318ms\n",
      "759:\tlearn: 0.2770005\ttotal: 1s\tremaining: 317ms\n",
      "760:\tlearn: 0.2768297\ttotal: 1s\tremaining: 315ms\n",
      "761:\tlearn: 0.2767614\ttotal: 1s\tremaining: 314ms\n",
      "762:\tlearn: 0.2766211\ttotal: 1.01s\tremaining: 313ms\n",
      "763:\tlearn: 0.2764537\ttotal: 1.01s\tremaining: 312ms\n",
      "764:\tlearn: 0.2763265\ttotal: 1.01s\tremaining: 310ms\n",
      "765:\tlearn: 0.2762628\ttotal: 1.01s\tremaining: 309ms\n",
      "766:\tlearn: 0.2761558\ttotal: 1.01s\tremaining: 307ms\n",
      "767:\tlearn: 0.2761449\ttotal: 1.01s\tremaining: 306ms\n",
      "768:\tlearn: 0.2760656\ttotal: 1.01s\tremaining: 305ms\n",
      "769:\tlearn: 0.2759638\ttotal: 1.01s\tremaining: 303ms\n",
      "770:\tlearn: 0.2758169\ttotal: 1.02s\tremaining: 302ms\n",
      "771:\tlearn: 0.2756715\ttotal: 1.02s\tremaining: 301ms\n",
      "772:\tlearn: 0.2754245\ttotal: 1.02s\tremaining: 299ms\n",
      "773:\tlearn: 0.2753669\ttotal: 1.02s\tremaining: 298ms\n",
      "774:\tlearn: 0.2752941\ttotal: 1.02s\tremaining: 296ms\n",
      "775:\tlearn: 0.2751908\ttotal: 1.02s\tremaining: 295ms\n",
      "776:\tlearn: 0.2750655\ttotal: 1.02s\tremaining: 294ms\n",
      "777:\tlearn: 0.2748963\ttotal: 1.02s\tremaining: 292ms\n",
      "778:\tlearn: 0.2748725\ttotal: 1.02s\tremaining: 291ms\n",
      "779:\tlearn: 0.2747756\ttotal: 1.02s\tremaining: 289ms\n",
      "780:\tlearn: 0.2747255\ttotal: 1.03s\tremaining: 288ms\n",
      "781:\tlearn: 0.2745880\ttotal: 1.03s\tremaining: 287ms\n",
      "782:\tlearn: 0.2745463\ttotal: 1.03s\tremaining: 285ms\n",
      "783:\tlearn: 0.2744945\ttotal: 1.03s\tremaining: 284ms\n",
      "784:\tlearn: 0.2743614\ttotal: 1.03s\tremaining: 283ms\n",
      "785:\tlearn: 0.2742093\ttotal: 1.03s\tremaining: 281ms\n",
      "786:\tlearn: 0.2741384\ttotal: 1.03s\tremaining: 280ms\n",
      "787:\tlearn: 0.2739845\ttotal: 1.03s\tremaining: 279ms\n",
      "788:\tlearn: 0.2739016\ttotal: 1.04s\tremaining: 277ms\n",
      "789:\tlearn: 0.2737033\ttotal: 1.04s\tremaining: 276ms\n",
      "790:\tlearn: 0.2736234\ttotal: 1.04s\tremaining: 275ms\n",
      "791:\tlearn: 0.2735290\ttotal: 1.04s\tremaining: 273ms\n",
      "792:\tlearn: 0.2734584\ttotal: 1.04s\tremaining: 272ms\n",
      "793:\tlearn: 0.2733755\ttotal: 1.04s\tremaining: 271ms\n",
      "794:\tlearn: 0.2732138\ttotal: 1.04s\tremaining: 269ms\n",
      "795:\tlearn: 0.2730944\ttotal: 1.04s\tremaining: 268ms\n",
      "796:\tlearn: 0.2730002\ttotal: 1.05s\tremaining: 267ms\n",
      "797:\tlearn: 0.2729169\ttotal: 1.05s\tremaining: 265ms\n",
      "798:\tlearn: 0.2727404\ttotal: 1.05s\tremaining: 264ms\n",
      "799:\tlearn: 0.2725824\ttotal: 1.05s\tremaining: 262ms\n",
      "800:\tlearn: 0.2725074\ttotal: 1.05s\tremaining: 261ms\n",
      "801:\tlearn: 0.2725023\ttotal: 1.05s\tremaining: 260ms\n",
      "802:\tlearn: 0.2723104\ttotal: 1.05s\tremaining: 258ms\n",
      "803:\tlearn: 0.2722122\ttotal: 1.05s\tremaining: 257ms\n",
      "804:\tlearn: 0.2719558\ttotal: 1.05s\tremaining: 255ms\n",
      "805:\tlearn: 0.2718889\ttotal: 1.05s\tremaining: 254ms\n",
      "806:\tlearn: 0.2718445\ttotal: 1.06s\tremaining: 253ms\n",
      "807:\tlearn: 0.2716387\ttotal: 1.06s\tremaining: 251ms\n",
      "808:\tlearn: 0.2715386\ttotal: 1.06s\tremaining: 250ms\n",
      "809:\tlearn: 0.2714043\ttotal: 1.06s\tremaining: 249ms\n",
      "810:\tlearn: 0.2713040\ttotal: 1.06s\tremaining: 247ms\n",
      "811:\tlearn: 0.2712522\ttotal: 1.06s\tremaining: 246ms\n",
      "812:\tlearn: 0.2710331\ttotal: 1.06s\tremaining: 245ms\n",
      "813:\tlearn: 0.2709458\ttotal: 1.06s\tremaining: 243ms\n",
      "814:\tlearn: 0.2709264\ttotal: 1.06s\tremaining: 242ms\n",
      "815:\tlearn: 0.2708176\ttotal: 1.07s\tremaining: 241ms\n",
      "816:\tlearn: 0.2706665\ttotal: 1.07s\tremaining: 239ms\n",
      "817:\tlearn: 0.2705544\ttotal: 1.07s\tremaining: 238ms\n",
      "818:\tlearn: 0.2704711\ttotal: 1.07s\tremaining: 236ms\n",
      "819:\tlearn: 0.2704014\ttotal: 1.07s\tremaining: 235ms\n",
      "820:\tlearn: 0.2703013\ttotal: 1.07s\tremaining: 234ms\n",
      "821:\tlearn: 0.2702386\ttotal: 1.07s\tremaining: 233ms\n",
      "822:\tlearn: 0.2700671\ttotal: 1.07s\tremaining: 231ms\n",
      "823:\tlearn: 0.2699848\ttotal: 1.08s\tremaining: 230ms\n",
      "824:\tlearn: 0.2699290\ttotal: 1.08s\tremaining: 229ms\n",
      "825:\tlearn: 0.2698131\ttotal: 1.08s\tremaining: 227ms\n",
      "826:\tlearn: 0.2696841\ttotal: 1.08s\tremaining: 226ms\n",
      "827:\tlearn: 0.2696131\ttotal: 1.08s\tremaining: 225ms\n",
      "828:\tlearn: 0.2694528\ttotal: 1.08s\tremaining: 223ms\n",
      "829:\tlearn: 0.2693714\ttotal: 1.08s\tremaining: 222ms\n",
      "830:\tlearn: 0.2692974\ttotal: 1.08s\tremaining: 220ms\n",
      "831:\tlearn: 0.2691344\ttotal: 1.08s\tremaining: 219ms\n",
      "832:\tlearn: 0.2691064\ttotal: 1.09s\tremaining: 218ms\n",
      "833:\tlearn: 0.2689752\ttotal: 1.09s\tremaining: 217ms\n",
      "834:\tlearn: 0.2688723\ttotal: 1.09s\tremaining: 215ms\n",
      "835:\tlearn: 0.2687154\ttotal: 1.09s\tremaining: 214ms\n",
      "836:\tlearn: 0.2685976\ttotal: 1.09s\tremaining: 212ms\n",
      "837:\tlearn: 0.2684540\ttotal: 1.09s\tremaining: 211ms\n",
      "838:\tlearn: 0.2683777\ttotal: 1.09s\tremaining: 210ms\n",
      "839:\tlearn: 0.2682344\ttotal: 1.09s\tremaining: 209ms\n",
      "840:\tlearn: 0.2681036\ttotal: 1.1s\tremaining: 207ms\n",
      "841:\tlearn: 0.2680310\ttotal: 1.1s\tremaining: 206ms\n",
      "842:\tlearn: 0.2679494\ttotal: 1.1s\tremaining: 205ms\n",
      "843:\tlearn: 0.2678957\ttotal: 1.1s\tremaining: 203ms\n",
      "844:\tlearn: 0.2678563\ttotal: 1.1s\tremaining: 202ms\n",
      "845:\tlearn: 0.2678279\ttotal: 1.1s\tremaining: 201ms\n",
      "846:\tlearn: 0.2676790\ttotal: 1.1s\tremaining: 199ms\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "847:\tlearn: 0.2676156\ttotal: 1.1s\tremaining: 198ms\n",
      "848:\tlearn: 0.2675713\ttotal: 1.1s\tremaining: 197ms\n",
      "849:\tlearn: 0.2675368\ttotal: 1.11s\tremaining: 195ms\n",
      "850:\tlearn: 0.2674520\ttotal: 1.11s\tremaining: 194ms\n",
      "851:\tlearn: 0.2674319\ttotal: 1.11s\tremaining: 193ms\n",
      "852:\tlearn: 0.2673353\ttotal: 1.11s\tremaining: 191ms\n",
      "853:\tlearn: 0.2671294\ttotal: 1.11s\tremaining: 190ms\n",
      "854:\tlearn: 0.2670178\ttotal: 1.11s\tremaining: 189ms\n",
      "855:\tlearn: 0.2669397\ttotal: 1.11s\tremaining: 187ms\n",
      "856:\tlearn: 0.2668316\ttotal: 1.11s\tremaining: 186ms\n",
      "857:\tlearn: 0.2667824\ttotal: 1.11s\tremaining: 184ms\n",
      "858:\tlearn: 0.2666784\ttotal: 1.11s\tremaining: 183ms\n",
      "859:\tlearn: 0.2665943\ttotal: 1.12s\tremaining: 182ms\n",
      "860:\tlearn: 0.2665080\ttotal: 1.12s\tremaining: 180ms\n",
      "861:\tlearn: 0.2664401\ttotal: 1.12s\tremaining: 179ms\n",
      "862:\tlearn: 0.2663388\ttotal: 1.12s\tremaining: 178ms\n",
      "863:\tlearn: 0.2662377\ttotal: 1.12s\tremaining: 176ms\n",
      "864:\tlearn: 0.2661083\ttotal: 1.12s\tremaining: 175ms\n",
      "865:\tlearn: 0.2659152\ttotal: 1.12s\tremaining: 174ms\n",
      "866:\tlearn: 0.2656316\ttotal: 1.12s\tremaining: 172ms\n",
      "867:\tlearn: 0.2655909\ttotal: 1.13s\tremaining: 171ms\n",
      "868:\tlearn: 0.2655672\ttotal: 1.13s\tremaining: 170ms\n",
      "869:\tlearn: 0.2654553\ttotal: 1.13s\tremaining: 168ms\n",
      "870:\tlearn: 0.2653773\ttotal: 1.13s\tremaining: 167ms\n",
      "871:\tlearn: 0.2652965\ttotal: 1.13s\tremaining: 166ms\n",
      "872:\tlearn: 0.2652204\ttotal: 1.13s\tremaining: 164ms\n",
      "873:\tlearn: 0.2651932\ttotal: 1.13s\tremaining: 163ms\n",
      "874:\tlearn: 0.2650540\ttotal: 1.13s\tremaining: 162ms\n",
      "875:\tlearn: 0.2649766\ttotal: 1.14s\tremaining: 161ms\n",
      "876:\tlearn: 0.2647709\ttotal: 1.14s\tremaining: 159ms\n",
      "877:\tlearn: 0.2646616\ttotal: 1.14s\tremaining: 158ms\n",
      "878:\tlearn: 0.2645373\ttotal: 1.14s\tremaining: 157ms\n",
      "879:\tlearn: 0.2643506\ttotal: 1.14s\tremaining: 156ms\n",
      "880:\tlearn: 0.2641617\ttotal: 1.14s\tremaining: 154ms\n",
      "881:\tlearn: 0.2641175\ttotal: 1.14s\tremaining: 153ms\n",
      "882:\tlearn: 0.2640663\ttotal: 1.15s\tremaining: 152ms\n",
      "883:\tlearn: 0.2639348\ttotal: 1.15s\tremaining: 150ms\n",
      "884:\tlearn: 0.2638803\ttotal: 1.15s\tremaining: 149ms\n",
      "885:\tlearn: 0.2638054\ttotal: 1.15s\tremaining: 148ms\n",
      "886:\tlearn: 0.2637431\ttotal: 1.15s\tremaining: 147ms\n",
      "887:\tlearn: 0.2636297\ttotal: 1.15s\tremaining: 145ms\n",
      "888:\tlearn: 0.2635628\ttotal: 1.15s\tremaining: 144ms\n",
      "889:\tlearn: 0.2633662\ttotal: 1.15s\tremaining: 143ms\n",
      "890:\tlearn: 0.2632123\ttotal: 1.15s\tremaining: 141ms\n",
      "891:\tlearn: 0.2631383\ttotal: 1.16s\tremaining: 140ms\n",
      "892:\tlearn: 0.2630376\ttotal: 1.16s\tremaining: 139ms\n",
      "893:\tlearn: 0.2630045\ttotal: 1.16s\tremaining: 137ms\n",
      "894:\tlearn: 0.2628480\ttotal: 1.16s\tremaining: 136ms\n",
      "895:\tlearn: 0.2627655\ttotal: 1.16s\tremaining: 135ms\n",
      "896:\tlearn: 0.2626969\ttotal: 1.16s\tremaining: 133ms\n",
      "897:\tlearn: 0.2626500\ttotal: 1.16s\tremaining: 132ms\n",
      "898:\tlearn: 0.2625846\ttotal: 1.16s\tremaining: 131ms\n",
      "899:\tlearn: 0.2625398\ttotal: 1.16s\tremaining: 129ms\n",
      "900:\tlearn: 0.2624811\ttotal: 1.17s\tremaining: 128ms\n",
      "901:\tlearn: 0.2624023\ttotal: 1.17s\tremaining: 127ms\n",
      "902:\tlearn: 0.2623097\ttotal: 1.17s\tremaining: 126ms\n",
      "903:\tlearn: 0.2621240\ttotal: 1.17s\tremaining: 124ms\n",
      "904:\tlearn: 0.2620605\ttotal: 1.17s\tremaining: 123ms\n",
      "905:\tlearn: 0.2618514\ttotal: 1.17s\tremaining: 122ms\n",
      "906:\tlearn: 0.2617891\ttotal: 1.17s\tremaining: 120ms\n",
      "907:\tlearn: 0.2616840\ttotal: 1.18s\tremaining: 119ms\n",
      "908:\tlearn: 0.2615402\ttotal: 1.18s\tremaining: 118ms\n",
      "909:\tlearn: 0.2613870\ttotal: 1.18s\tremaining: 116ms\n",
      "910:\tlearn: 0.2613267\ttotal: 1.18s\tremaining: 115ms\n",
      "911:\tlearn: 0.2612668\ttotal: 1.18s\tremaining: 114ms\n",
      "912:\tlearn: 0.2612220\ttotal: 1.18s\tremaining: 112ms\n",
      "913:\tlearn: 0.2610641\ttotal: 1.18s\tremaining: 111ms\n",
      "914:\tlearn: 0.2608815\ttotal: 1.18s\tremaining: 110ms\n",
      "915:\tlearn: 0.2608238\ttotal: 1.18s\tremaining: 109ms\n",
      "916:\tlearn: 0.2607264\ttotal: 1.19s\tremaining: 107ms\n",
      "917:\tlearn: 0.2606168\ttotal: 1.19s\tremaining: 106ms\n",
      "918:\tlearn: 0.2605380\ttotal: 1.19s\tremaining: 105ms\n",
      "919:\tlearn: 0.2605049\ttotal: 1.19s\tremaining: 103ms\n",
      "920:\tlearn: 0.2603490\ttotal: 1.19s\tremaining: 102ms\n",
      "921:\tlearn: 0.2602846\ttotal: 1.19s\tremaining: 101ms\n",
      "922:\tlearn: 0.2602087\ttotal: 1.19s\tremaining: 99.5ms\n",
      "923:\tlearn: 0.2601822\ttotal: 1.19s\tremaining: 98.2ms\n",
      "924:\tlearn: 0.2600402\ttotal: 1.19s\tremaining: 96.9ms\n",
      "925:\tlearn: 0.2599298\ttotal: 1.2s\tremaining: 95.6ms\n",
      "926:\tlearn: 0.2598606\ttotal: 1.2s\tremaining: 94.3ms\n",
      "927:\tlearn: 0.2597281\ttotal: 1.2s\tremaining: 93ms\n",
      "928:\tlearn: 0.2596532\ttotal: 1.2s\tremaining: 91.7ms\n",
      "929:\tlearn: 0.2595868\ttotal: 1.2s\tremaining: 90.3ms\n",
      "930:\tlearn: 0.2594280\ttotal: 1.2s\tremaining: 89ms\n",
      "931:\tlearn: 0.2593269\ttotal: 1.2s\tremaining: 87.7ms\n",
      "932:\tlearn: 0.2591151\ttotal: 1.2s\tremaining: 86.4ms\n",
      "933:\tlearn: 0.2590226\ttotal: 1.2s\tremaining: 85.1ms\n",
      "934:\tlearn: 0.2588045\ttotal: 1.21s\tremaining: 83.8ms\n",
      "935:\tlearn: 0.2587398\ttotal: 1.21s\tremaining: 82.5ms\n",
      "936:\tlearn: 0.2586917\ttotal: 1.21s\tremaining: 81.2ms\n",
      "937:\tlearn: 0.2586307\ttotal: 1.21s\tremaining: 79.9ms\n",
      "938:\tlearn: 0.2586082\ttotal: 1.21s\tremaining: 78.6ms\n",
      "939:\tlearn: 0.2585072\ttotal: 1.21s\tremaining: 77.3ms\n",
      "940:\tlearn: 0.2583942\ttotal: 1.21s\tremaining: 76ms\n",
      "941:\tlearn: 0.2583132\ttotal: 1.21s\tremaining: 74.7ms\n",
      "942:\tlearn: 0.2582571\ttotal: 1.21s\tremaining: 73.4ms\n",
      "943:\tlearn: 0.2582059\ttotal: 1.22s\tremaining: 72.1ms\n",
      "944:\tlearn: 0.2579891\ttotal: 1.22s\tremaining: 70.8ms\n",
      "945:\tlearn: 0.2579483\ttotal: 1.22s\tremaining: 69.5ms\n",
      "946:\tlearn: 0.2578234\ttotal: 1.22s\tremaining: 68.2ms\n",
      "947:\tlearn: 0.2578027\ttotal: 1.22s\tremaining: 66.9ms\n",
      "948:\tlearn: 0.2577104\ttotal: 1.22s\tremaining: 65.6ms\n",
      "949:\tlearn: 0.2575642\ttotal: 1.22s\tremaining: 64.3ms\n",
      "950:\tlearn: 0.2575561\ttotal: 1.22s\tremaining: 63ms\n",
      "951:\tlearn: 0.2572783\ttotal: 1.22s\tremaining: 61.7ms\n",
      "952:\tlearn: 0.2572253\ttotal: 1.22s\tremaining: 60.4ms\n",
      "953:\tlearn: 0.2571799\ttotal: 1.23s\tremaining: 59.1ms\n",
      "954:\tlearn: 0.2571248\ttotal: 1.23s\tremaining: 57.8ms\n",
      "955:\tlearn: 0.2570646\ttotal: 1.23s\tremaining: 56.5ms\n",
      "956:\tlearn: 0.2570143\ttotal: 1.23s\tremaining: 55.2ms\n",
      "957:\tlearn: 0.2568050\ttotal: 1.23s\tremaining: 53.9ms\n",
      "958:\tlearn: 0.2567678\ttotal: 1.23s\tremaining: 52.6ms\n",
      "959:\tlearn: 0.2567020\ttotal: 1.23s\tremaining: 51.3ms\n",
      "960:\tlearn: 0.2564308\ttotal: 1.23s\tremaining: 50ms\n",
      "961:\tlearn: 0.2564107\ttotal: 1.23s\tremaining: 48.7ms\n",
      "962:\tlearn: 0.2562693\ttotal: 1.23s\tremaining: 47.4ms\n",
      "963:\tlearn: 0.2562285\ttotal: 1.24s\tremaining: 46.1ms\n",
      "964:\tlearn: 0.2561045\ttotal: 1.24s\tremaining: 44.8ms\n",
      "965:\tlearn: 0.2559232\ttotal: 1.24s\tremaining: 43.6ms\n",
      "966:\tlearn: 0.2557107\ttotal: 1.24s\tremaining: 42.3ms\n",
      "967:\tlearn: 0.2556461\ttotal: 1.24s\tremaining: 41ms\n",
      "968:\tlearn: 0.2555025\ttotal: 1.24s\tremaining: 39.7ms\n",
      "969:\tlearn: 0.2554092\ttotal: 1.24s\tremaining: 38.4ms\n",
      "970:\tlearn: 0.2552257\ttotal: 1.24s\tremaining: 37.1ms\n",
      "971:\tlearn: 0.2551102\ttotal: 1.24s\tremaining: 35.8ms\n",
      "972:\tlearn: 0.2550249\ttotal: 1.24s\tremaining: 34.5ms\n",
      "973:\tlearn: 0.2550073\ttotal: 1.25s\tremaining: 33.3ms\n",
      "974:\tlearn: 0.2549131\ttotal: 1.25s\tremaining: 32ms\n",
      "975:\tlearn: 0.2548554\ttotal: 1.25s\tremaining: 30.7ms\n",
      "976:\tlearn: 0.2547860\ttotal: 1.25s\tremaining: 29.4ms\n",
      "977:\tlearn: 0.2547009\ttotal: 1.25s\tremaining: 28.1ms\n",
      "978:\tlearn: 0.2546768\ttotal: 1.25s\tremaining: 26.8ms\n",
      "979:\tlearn: 0.2544005\ttotal: 1.25s\tremaining: 25.6ms\n",
      "980:\tlearn: 0.2543705\ttotal: 1.25s\tremaining: 24.3ms\n",
      "981:\tlearn: 0.2542915\ttotal: 1.25s\tremaining: 23ms\n",
      "982:\tlearn: 0.2542708\ttotal: 1.25s\tremaining: 21.7ms\n",
      "983:\tlearn: 0.2542098\ttotal: 1.26s\tremaining: 20.4ms\n",
      "984:\tlearn: 0.2540815\ttotal: 1.26s\tremaining: 19.2ms\n",
      "985:\tlearn: 0.2537973\ttotal: 1.26s\tremaining: 17.9ms\n",
      "986:\tlearn: 0.2536962\ttotal: 1.26s\tremaining: 16.6ms\n",
      "987:\tlearn: 0.2535510\ttotal: 1.26s\tremaining: 15.3ms\n",
      "988:\tlearn: 0.2534875\ttotal: 1.26s\tremaining: 14ms\n",
      "989:\tlearn: 0.2533538\ttotal: 1.26s\tremaining: 12.8ms\n",
      "990:\tlearn: 0.2532907\ttotal: 1.26s\tremaining: 11.5ms\n",
      "991:\tlearn: 0.2530947\ttotal: 1.26s\tremaining: 10.2ms\n",
      "992:\tlearn: 0.2528559\ttotal: 1.27s\tremaining: 8.93ms\n",
      "993:\tlearn: 0.2527918\ttotal: 1.27s\tremaining: 7.65ms\n",
      "994:\tlearn: 0.2527223\ttotal: 1.27s\tremaining: 6.38ms\n",
      "995:\tlearn: 0.2526244\ttotal: 1.27s\tremaining: 5.1ms\n",
      "996:\tlearn: 0.2524156\ttotal: 1.27s\tremaining: 3.83ms\n",
      "997:\tlearn: 0.2523993\ttotal: 1.27s\tremaining: 2.55ms\n",
      "998:\tlearn: 0.2522552\ttotal: 1.27s\tremaining: 1.27ms\n",
      "999:\tlearn: 0.2522183\ttotal: 1.27s\tremaining: 0us\n",
      "Test Sonuçları:\n",
      "\n",
      "accuracy score: 0.8156\n",
      "\n",
      "Classification Report: \n",
      "               precision    recall  f1-score   support\n",
      "\n",
      "           0       0.79      0.93      0.86       106\n",
      "           1       0.87      0.64      0.74        73\n",
      "\n",
      "    accuracy                           0.82       179\n",
      "   macro avg       0.83      0.79      0.80       179\n",
      "weighted avg       0.82      0.82      0.81       179\n",
      "\n",
      "\n",
      "Confusion Matrix: \n",
      " [[99  7]\n",
      " [26 47]]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from catboost import CatBoostClassifier\n",
    "model=CatBoostClassifier()\n",
    "model.fit(X_train,y_train)\n",
    "acc_cat = model.score(X_test,y_test)\n",
    "print_score(model,X_train, y_train, X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "importances = pd.DataFrame({'feature':X_train.columns,'importance':np.round(model.feature_importances_,3)})\n",
    "importances = importances.sort_values('importance',ascending=False).set_index('feature')\n",
    "importances.plot.bar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Score</th>\n",
       "      <th>Model</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.832402</td>\n",
       "      <td>GBM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.815642</td>\n",
       "      <td>CatBoost</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.810056</td>\n",
       "      <td>XGBoost</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.793296</td>\n",
       "      <td>Random Forest</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.787709</td>\n",
       "      <td>LGBM</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.754190</td>\n",
       "      <td>Bagging-Meta Estimator</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Score                   Model\n",
       "0  0.832402                     GBM\n",
       "1  0.815642                CatBoost\n",
       "2  0.810056                 XGBoost\n",
       "3  0.793296           Random Forest\n",
       "4  0.787709                    LGBM\n",
       "5  0.754190  Bagging-Meta Estimator"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "results = pd.DataFrame({\n",
    "    'Model': ['Bagging-Meta Estimator', 'Random Forest', 'GBM', \n",
    "              'XGBoost','LGBM','CatBoost'],\n",
    "    'Score': [acc_bag, acc_rf,acc_gbm,acc_xgb,acc_lgb,acc_cat]})\n",
    "result_df = results.sort_values(by='Score', ascending=False)\n",
    "result_df = result_df.set_index('Score')\n",
    "result_df.reset_index(inplace=True)\n",
    "result_df.head(9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import seaborn as sns\n",
    "sns.barplot(x='Model',y='Score',data =result_df)\n",
    "plt.ylim(0.75,0.85)\n",
    "plt.xticks(rotation=20)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Sonuç:\n",
    "Çeşitli topluluk öğrenme tekniklerini ele aldık ve bu tekniklerin makine öğrenimi algoritmalarında nasıl uygulandığını gördük. Ayrıca, titanic veri setimize ön işleme,kayıp veri analizi yaparak ardından bu algoritmaları uyguladık. Ekrandaki bar grafiğinde hangi modelin test seti üzerinde daha başarılı olduğunu gözlemleyebiliriz. Bu durumda GBM modelimiz en başarılı modeldir. XGBoost LightGBM gibi algoritmaların GBM'in büyük veri setlerindeki yavaşlığı ve düşük accuracy skoru yüzünden geliştirildiğini biliyoruz. Bu durumda bizim veri setimizdeki gözlem sayısı az olduğu için GBM algoritmasının daha başarılı çıkması da normaldir.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Kaynaklar\n",
    "\n",
    "https://medium.com/sfu-cspmp/xgboost-a-deep-dive-into-boosting-f06c9c41349\n",
    "https://subscription.packtpub.com/book/big_data_and_business_intelligence/9781789136609/2/ch02lvl1sec16/max-voting\n",
    "https://xgboost.readthedocs.io/en/latest/R-package/discoverYourData.html\n",
    "https://www.analyticsvidhya.com/blog/2018/06/comprehensive-guide-for-ensemble-models/\n",
    "https://www.veribilimiokulu.com/blog/gradient-boosted-regresyon-agaclari/\n",
    "https://www.veribilimiokulu.com/blog/xgboost-nasil-calisir/\n",
    "https://www.analyticsvidhya.com/blog/2018/09/an-end-to-end-guide-to-understand-the-math-behind-xgboost/\n",
    "https://medium.com/deep-learning-turkiye/xgboost-dosyalar%C4%B1-25d31fda6900"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}