Bank-Loan-Defaulter-Prediction

🔗 Project Files and Links

• 💻 App Link - Interact with the model in real-time.
• 💻 Notebook Link - Explore the project.

Introduction

This project focuses on predicting loan defaulters using a dataset provided by MachineHack for the Deloitte Hackathon. With the increasing number of bad loans impacting banks' profitability and the country's economy, accurate prediction of loan defaults has become crucial.

Goal

The objective is to develop a machine learning model that can predict whether a customer will default on their loan based on various financial and demographic attributes, such as loan amount and interest rate.

Key Highlights

• Built a predictive model using Logistic Regression, Decision Tree, Random Forest, Gradient Boosting, and XGBoost, with the final Hyperparameter Tuned XGBoost model achieving ~93% accuracy and 91% recall using SMOTE.
• Conducted EDA and feature engineering for model building and built a Streamlit app for loan defaulter prediction.
• Delivered insights to reduce default risks and optimize loan approval strategies.