Prediction of Personal Credit Rates with Incomplete Data Sets Using Cognitive Mapping

This study suggests a Naive Bayesian style method called Frequency Matrix technique, which simulates cognitive mapping in human brain, to predict personal credit rates with incomplete data sets. Its performance is compared with that of multiple discriminant analysis and logistic regression. Missing values are predicted with mean imputation method and regression imputation method for these two methods. An artificial neural network is also introduced and tested for their performance. A data set on personal credit information of 8,234 customers of Bank A is collected for the tests. The performance of Frequency Matrix technique is compared with that of other methods. The results from the tests show that the performance of Frequency Matrix technique is superior to that of other methods such as MDA-mean, Logit-mean, MDA-regression, Logit-regression, and Artificial Neural Networks.