Session 4: Rprofet – A Library for Fast/Stable Credit Scorecards
Presentation Type
Event
Abstract
Credit scoring offers an objective way to make lending decisions based on industry experience. The goal of credit scoring models such as credit risk scorecards, is to provide predictive information about the potential default of customers and to minimize the risk of loss by setting approval decision rules. Logistic regression is commonly used to develop credit scoring models when the target is a binary qualitative variable. The Rprofet package performs various steps in the credit scoring process and allows users to customize each step to build an effective scorecard using logistic regression. The package provides functions for coarse variable binning, Weights of Evidence (WOE) transformation, variable clustering, custom binning, visualization, and scaling of scorecards. The final scorecard can provide consistent and reasonable credit decisions based on user’s definition of creditworthiness.
Start Date
2-6-2024 11:00 AM
End Date
2-6-2024 12:00 PM
Session 4: Rprofet – A Library for Fast/Stable Credit Scorecards
Herold Crest 253 C
Credit scoring offers an objective way to make lending decisions based on industry experience. The goal of credit scoring models such as credit risk scorecards, is to provide predictive information about the potential default of customers and to minimize the risk of loss by setting approval decision rules. Logistic regression is commonly used to develop credit scoring models when the target is a binary qualitative variable. The Rprofet package performs various steps in the credit scoring process and allows users to customize each step to build an effective scorecard using logistic regression. The package provides functions for coarse variable binning, Weights of Evidence (WOE) transformation, variable clustering, custom binning, visualization, and scaling of scorecards. The final scorecard can provide consistent and reasonable credit decisions based on user’s definition of creditworthiness.