Supervised Learning Model test

This repository contains a simple implementation of a logistic regression and RandomForestClassifier model for predicting diabetes based on various features. The model is trained on a dataset containing information related to diabetes symptoms.

The dataset used for training the model includes the following features:

['Age', 'Gender', 'Polyuria', 'Polydipsia', 'sudden weight loss', 'weakness', 'Polyphagia', 'Genital thrush', 'visual blurring', 'Itching', 'Irritability', 'delayed healing', 'partial paresis', 'muscle stiffness', 'Alopecia', 'Obesity', 'class']

The target variable is the 'class', indicating the presence or absence of diabetes.

I started by training the RandomForestClassifier model on the data based on the available features. A classification report and confusion matrix was done. The result was this:

I now used feature importance to select the best features and trained another RandomForestClassifier Model on the data now based on the selected features (these features were selected on a threshold of >0.05). A classification report and confusion matrix was done on it, below is the result:

The next method i implemented was SelectFromModel. I used this on the data set and RandomForestClassifier. SelectFromModel is used to automatically select features based on their importance scores with a specified threshold. A classification report, the features selected and confusion matrix was done. Below is the result:

These same processes were done with LogisticRegression

The complete analysis can be found in the notebook in this repository