P2P Loan Default Prediction and Risk Estimation System built using Machine Learning Techniques – Python
• Consulted an external risk management consultancy firm to provide data on loan default status of several customers and applied various data cleaning tools and used that data to structure a Machine Learning algorithm. • Predicted default & risk probability using GNB, Lasso regression, SVM and Deep Learning, achieving 91% accuracy. • Presented findings to the company and submitted a detailed PDF report for the same.