Development of a Machine Learning Model - Loan Repayment Prediction

Problem Statement 

The project aims to solve the problem of assessing the likelihood of loan repayment to assist lenders in making informed decisions regarding loan approvals and managing risk effectively.

Approach:

The problem was addressed by developing a predictive model using an artificial neural network. The approach involved:

• Cleaning and preprocessing raw data using Numpy and Pandas to ensure data quality.
• Applying one-hot encoding to handle categorical variables and performing feature engineering to create meaningful predictors.
• Designing and training the neural network with ReLU and Sigmoid activation functions and the Adam optimizer for accurate predictions.

The model provided accurate predictions that effectively supported decision-making for loan approvals and risk management. Its performance was measured using evaluation metrics, demonstrating its reliability and effectiveness in solving the problem