EMI Bounce Predictor

Table of Contents

• About
• Problem Statement
• Project Workflow
• Data Pre-Processing
• Model Building (Training, Testing & Hyperparameter Optimization)
• Model Evaluation
• Model Explanation using SHAP (SHapley Additive exPlanations)
• Deployment

About

• Built a machine learning model to predict which clients are most likely to default on their EMI payments for tractor loans.
• Preprocessed data, performed EDA, trained and tested various ML models and deployed a web application based on the model on Heroku.
• This project was submitted for TVS Credit E.P.I.C. Analytics Challenge Season 4 (2022).

Problem Statement

Project Workflow

Data Pre-Processing

Model Building (Training, Testing & Hyperparameter Optimization)

Model Evaluation

Model Explanation using SHAP (SHapley Additive exPlanations)

Deployment