Credit Risk Classification Using Kernel Logistic Regression-least Square Support Vector Machine

Kernel Logistic Regression (KLR) is one of the statistical models that have been proposed for classification in the machine learning and data mining communities, and also one of the effective methodologies in the kernel-machine techniques. The parameters of KLR model are usually fitted by the solution of a convex optimization problem that can be found using the well known Iteratively Reweighted Least Squares (IRLS) algorithm. In this research, we use the Least Squares Support Vector Machine (LS-SVM) framework to solve the KLR problem. Recently, KLR has become very popular because it allows non-linear probabilistic classification which provides competitive discriminative ability and transparent reasoning to classify, also can be generalized to multi-class problem. Credit risk data sets from UCI machine learning are used in order to verify the effectiveness of the KLR-IRLSSVM method to evaluate credit risk in bank in a reasonable way.