Overview

This project, is centred around the development of predictive models for hospital readmission, particularly targeting diabetic patients due to their substantial impact on healthcare expenses. The primary hypothesis posits that machine learning algorithms can effectively predict readmissions, thus contributing to cost reduction and enhanced patient care.

The study entails two predictive tasks: binary classification to ascertain the likelihood of a patient's readmission within 30 days post-discharge, and multiclass classification to predict the readmission timeframe, categorized as 'No', '<30 days', or '>30 days'.

Full Report

The full Report, with detailed explanations, is present in the 'Machine_Learning_Group35_Report.pdf' file.