Machine-Learning

Machine Learning

Setup:

conda create -n ml conda activate ml

##1.1 Gather the data and build cross-validation

• Download this dataset: Categorical Feature Encoding Challenge from https://www.kaggle.com/c/cat-in-the-dat/data then save it into input folder

• Check csv file: head train.csv

• python .\src\create_folds.py

• python .\src\train.py Result: 0.7437763648654564

##1.2 Building an inference for the machine learning framework run.bat set TRAINING_DATA=input\train_folds.csv set TEST_DATA=input\test.csv

set MODEL=%1

set FOLD=0 
python -m src.train
set FOLD=1 
python -m src.train
set FOLD=2 
python -m src.train
set FOLD=3 
python -m src.train
set FOLD=4 
python -m src.train

python -m src.predict

.\run.bat randomforest

2. Cross Validation Framework

• dataset for binary_classification and multi class classification same as above
• dataset for single_col_classification is https://www.kaggle.com/c/house-prices-advanced-regression-techniques/data

3. Handling categorical features in ML problems

• Dataset : https://www.kaggle.com/c/cat-in-the-dat-ii

• count number rows csv: wc -l ../input/train_cat.csv (linux only)

Ref:

run.bat in window set TRAINING_DATA=input\train_folds.csv set FOLD=0 python -m src.train1

and run.sh linux: set ==> export