Categorical Principal Component Logistic Regression: A Case Study for Housing Loan Approval

Abstract

The logistic regression describes the relationship between a binary (dichotomous) response variable and explanatory variables. If there is multi collinearity among the explanatory variables, the estimation of model parameters may lead to invalid statistical inference. In this study, we have survey data for 2331 randomly selected customers which consists of highly correlated binary explanatory variables to model whether a customer's housing loan application has been approved or not. For this purpose, we present a categorical principal component analysis to deal with the multi collinearity problem among categorical explanatory variables while predicting binary response variable with logistic regression. (C) 2014 The Authors. Published by Elsevier Ltd.