Liver Disease Prediction using Machine Learning

Project Overview

Machine learning system to predict liver disease from blood test results. Compares multiple algorithms using the Indian Liver Patient Dataset.

Results

• Best Model: Logistic Regression
• Accuracy: 73.5%
• Recall: 96.4%
• Top Features: ALT (Sgpt), AST (Sgot), Direct Bilirubin

Project Structure

• liver_disease_analysis.ipynb - Complete analysis notebook
• src/ - Data loading and model training code
• models/ - Trained model and scaler
• reports/ - Visualizations and results

Quick Start

pip install -r requirements.txt
python -c "from src.model_trainer import train_and_evaluate_models; results = train_and_evaluate_models()"