Score Engineered Logistic Regression

TL;DR

This paper introduces a sequential quadratic programming algorithm for score engineered logistic regression, enabling complex score shaping with constraints, and enhancing its application in credit scoring models.

Contribution

It develops a novel optimization method based on quadratic programming and Taylor series expansion to improve logistic regression score engineering.

Findings

Enables complex score shaping with constraints

Improves logistic regression performance for scoring models

Provides a new optimization algorithm for score engineering

Abstract

In several FICO studies logistic regression has been shown to be a very competitive technology for developing unrestricted scoring models, especially for performance metrics like ROC area. Application of logistic regression has been hampered by the lack of software to handle complex score engineering such as shape and pattern constraints. The purpose of this paper is to develop a sequential quadratic programming algorithm for score engineered logistic regression. This approach is based on a simple Taylor series expansion of minus log likelihood, which is locally quadratic. and fits in with the method that applies quadratic programming to B-Splines.