Dorota Witkowska, Ph.D., Dept. of Econometrics & Statistics, Warsaw University of Life Sciences, ul. Nowoursynowska 166, Warszawa, 02-787, Poland and Mariola Chrzanowska, Master, of, Econ, University of Economics and Administration, : ul. Karczówkowska 41, Kielce, 25-713, Poland.
The aim of the paper is dichotomous classification of the individual borrowers to the groups of creditworthy or non-creditworthy clients. The research is conducted on the basis of actual data regarding the individual borrowers who got mortgage credits in one of the commercial banks that operates in Poland. Each of the 2547 clients is described by features that are both quantitative and qualitative. Employing data, concerning all characteristics of the borrowers, we construct 11 variables that are used in classification procedures. The grouping variable informs if the client pays off the credit regularly due to the credit agreement or he is back in loan redemption. Diagnostic variables describe the clients in terms of demographic features (as: gender, place of living, etc.) and characterize the credits that are to be paid back (i.e. value and currency of the credit, credit rate, duration of the credit, etc.).In the paper the results of application selected methods to the individual credit repayment are presented. These methods are: binary regression, Bayesian discriminant analysis and aggregated classification trees (i.e. employing boosting and bagging algorithms). The efficiency of applied methods is evaluated in terms of chosen measures (i.e. general classification errors and accuracy in recognition the non-creditworthy clients).