Challenger Models in Credit Scoring - Publication
Logistic Regression versus state-of-the-art Data Mining Techniques
Status: Under revision.
Dr. Roman Timofeev (Deutsche Bank AG), Maximilian Hahn (Enrion GmbH) im Rahmen des SIP
Retail credit scoring is a typical area of application for data analytics and statistical forecasting. Recent developments in the areas of machine learning and artificial intelligence vastly improved the performance of predictive models. This article investigates the performance and feasibility of several state-of-the-art techniques on a real life retail client data set from a large financial institution.