supervised-learning's Introduction

Census-Income-Classification

This is a classification model to predict whether a person have more than 50,000 dollars salary.

Setup: python3.6 pandas numpy sklearn

(1)training process Usage: python train_classify.py -i -c -f <selection_model> -o

 Example:
     python train_classify.py -i data/census-income.data  -c boot -f forest -o forest_model.sav

 Options:
     -i: input file 
     -c: classifier methods: kfold or boot
     -f: classification model: logistic, forest, boosting
     -o: output file of training model

(2) prediction process

 Usage:
     python predict_classify.py -i <inputfile> -m <model file> 

 Example:
     python predict_classify.py -i data/classification-test.data -m forest_model.sav 

 Options:
     -i: input file 
     -m: model file

This is a selection method for significant features. Setup: python3.6 pandas numpy sklearn

Usage: python feature_FSS.py -i -f <feature_selection_model>

Example: python feature_FSS.py -i data/census-income.data -f forest

Options: -i: input file -f: feature selection model: forward, forest, boosting

This is a segementation model to create data segments.

Setup: python3.6 pandas numpy sklearn

(1) training process

 Usage:
     python train_seg.py -i <inputfile> -cluster <segementation_model> -n <number of clusters> -o <model output>

 Example:
     python train_seg.py -i data/census-income.data -cluster KNN -n 10 -o seg_model.sav

 Options:
     -i: input file 
     -cluster: segementation model: KNN or GMM
     -o: output file of training model

(2) prediction process

 Usage:
     python predict_seg.py -i <inputfile> -m <model file> 
 
 Example:
     python predict_seg.py -i data/segmentation-test.data -m seg_model.sav

 Options:
     -i: input file 
     -m: model file

Generating feature histogram for analysis Usage: python features_analysis.py -i -o

Example: python features_analysis.py -i data/census-income.data -o features_analysis.txt

Options: -i: input file -o: output file