artificial_intelligence/qsvm_variational.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## _*Quantum SVM (variational method)*_\n",
    "\n",
    "The QSVM_Kernel notebook here demonstrates a kernel based approach. This notebook shows a variational method.\n",
    "\n",
    "For further information please see: [https://arxiv.org/pdf/1804.11326.pdf](https://arxiv.org/pdf/1804.11326.pdf)\n",
    "\n",
    "\n",
    "**This notebook shows the SVM implementation based on the variational method.**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from datasets import *\n",
    "from qiskit_aqua.utils import split_dataset_to_data_and_labels, map_label_to_class_name\n",
    "from qiskit_aqua.input import get_input_instance\n",
    "from qiskit_aqua import run_algorithm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First we prepare the dataset, which is used for training, testing and the finally prediction.\n",
    "\n",
    "*Note: You can easily switch to a different dataset, such as the Breast Cancer dataset, by replacing 'ad_hoc_data' to 'Breast_cancer' below.*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fa77a2142b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fa77a104518>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'A': 0, 'B': 1}\n"
     ]
    }
   ],
   "source": [
    "n = 2 # dimension of each data point\n",
    "training_dataset_size = 20\n",
    "testing_dataset_size = 10\n",
    "\n",
    "sample_Total, training_input, test_input, class_labels = ad_hoc_data(training_size=training_dataset_size, \n",
    "                                                                     test_size=testing_dataset_size, \n",
    "                                                                     n=n, gap=0.3, PLOT_DATA=True)\n",
    "\n",
    "datapoints, class_to_label = split_dataset_to_data_and_labels(test_input)\n",
    "print(class_to_label)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With the dataset ready we initialize the necessary inputs for the algorithm:\n",
    "- the input dictionary (params) \n",
    "- the input object containing the dataset info (algo_input)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "params = {\n",
    "    'problem': {'name': 'svm_classification', 'random_seed': 10598},\n",
    "    'algorithm': {'name': 'QSVM.Variational', 'override_SPSA_params': True},\n",
    "    'backend': {'name': 'qasm_simulator', 'shots': 1024},\n",
    "    'optimizer': {'name': 'SPSA', 'max_trials': 200, 'save_steps': 1},\n",
    "    'variational_form': {'name': 'RYRZ', 'depth': 3},\n",
    "    'feature_map': {'name': 'SecondOrderExpansion', 'depth': 2}\n",
    "}\n",
    "\n",
    "algo_input = get_input_instance('SVMInput')\n",
    "algo_input.training_dataset  = training_input\n",
    "algo_input.test_dataset = test_input\n",
    "algo_input.datapoints = datapoints[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "With everything setup, we can now run the algorithm.\n",
    "\n",
    "For the testing, the result includes the details and the success ratio.\n",
    "\n",
    "For the prediction, the result includes the predicted labels. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "testing success ratio:  1.0\n",
      "predicted classes: ['A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'A', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B', 'B']\n"
     ]
    }
   ],
   "source": [
    "result = run_algorithm(params, algo_input)\n",
    "print(\"testing success ratio: \", result['testing_accuracy'])\n",
    "print(\"predicted classes:\", result['predicted_classes'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}