Datanalysis Credit Risk

A comprehensive credit risk preprocessing pipeline that handles the messy reality of financial data before you can even think about modeling. It runs 11 sequential steps including IV filtering, PSI stability checks, null importance denoising, and correlation removal, then spits out a detailed Excel report with 15 sheets covering everything from missing value patterns to organization-level statistics. Built for pre-loan modeling scenarios where you need to separate signal from noise across multiple lending organizations and time periods. The modular design lets you run steps independently, which is clutch when stakeholders want to see exactly why certain variables got axed.