Day 2 - Tech Talks 29 May
How to develop Credit risk models (scorecard) using Machine learning technique and its use in underwriting strategy development.
The Banking and finance domain is the early adopter of data analytics. Risk management is a million-dollar problem and the industry has evolved to use the sophisticated model to predict the bad loans, fraudulent transactions, etc. Credit risk modeling is ever-changing due to the recent popularity of ML and AI and data availability from various digital and alternate data sources.
The challenge faced by the data scientist is to improve the accuracy using sophisticated supervised learning algorithms while following the guideline of the central bank from a regulatory perspective. In the session, we would learn in-depth step by step methods to develop credit score, its use, validation of the models, and recalibration techniques.