Loan Default PredictionPredict which borrowers are likely to default on their loans by building an end-to-end pipeline—including data sampling, preprocessing, feature engineering, model training, and deployment with a Gradio UI—and deploy it as a live app on Hugging Face Spaces.
Agenda for the session
- Understand default risk, NPL targets, and use the full 5,000-row dataset
- Clean data by dropping IDs and one-hot encoding categoricals
- Train a Random Forest, evaluate with Accuracy & ROC-AUC, and save with joblib
- Build a Gradio app and deploy on Hugging Face with five project files
About Speakers
Dr. Davood Wadi
AI Research Scientist - intelChain
Dr. Davood Wadi is an AI Research Scientist at intelChain. Before joining intelChain, he excelled as an AI researcher and pursued his Ph.D. at HEC Montreal, renowned globally for its academic excellence. His interest in applying modern technologies to data sparked his tenure as a financial analyst, where he started incorporating mathematical methods into mass psychology to understand investment patterns. Davood’s expertise and interests include developing new algorithms for AI and ML applications, computer vision, NLP, Meta-Learning, and Consumer neuroscience.
