Using Data Mining Techniques to Detect Customer Default Payment

With the rapid growth of the financial market, the management of credit card recipients by financial institute is becoming increasingly significant. In order to retain the valuable customers and prevent churn by providing the right incentives, analysis of customer default payment data are required. A Linear Discrimination Regression (LDA) model is used to analysis of the data of customer default rate in Taiwan. The results of final regression model were shown using scorecard. Discussion of the KS Test results determines improvement in performance that financial institute will get from this risk model.