Credit Scoring Using Logistic Regression - An Auto Loan Case

This project involves the analysis of a dataset of auto-finance consumer data, including credit–bureau information, loan-application data, and company data regarding consumers’ performance since acquisition. The goal of this project was to analyze the dataset and create a prediction model to evaluate the credit-worthiness of new loan applicants. A logistic regression model was applied and the scorecard cut-off was set up.

Project Site (including report, powerpoint slides, data, sas code and other related materials)

Note: the goodsite service provides a good documentation tool to compile this project. I will update this project with more value-added stuff such as decision tree model, neutural network, and discriminant analysis.