End-to-end AI/ML projects, from data preprocessing to deployment
Predicts loan repayment risk using financial history. Built with LightGBM, Flask, and Angular.
This project is based on the Home Credit Default Risk Kaggle competition dataset and is not affiliated with or endorsed by Home Credit Group.
It aims to predict the probability of a loan applicant defaulting on a loan based on a wide range of historical financial, demographic, and behavioral data. The dataset includes alternative data sources such as telco usage and transactional information to improve credit risk assessment.
The model was trained using seven datasets from the competition, which were individually processed, aggregated, and merged to create a comprehensive feature set. A LightGBM model was then trained and deployed as a web app, allowing users to test predictions in real-time.