SDÜ İktisadi ve İdari Bilimler Fakültesi Dergisi, Cilt 14, Sayı 2 (2009)

CLASSIFICATION OF DOMESTIC AND FOREIGN COMMERCIAL BANKS IN TURKEY BASED ON FINANCIAL EFFICIENCY: A COMPARISON OF DECISION TREE, LOGISTIC REGRESSION AND DISCRIMINANT ANALYSIS MODELS

Ali Sait ALBAYRAK

Özet


This study compares the data mining (DM) techniques of linear dis-criminant analysis (LD), logistic regression (LR) and classification and re-gression tree analysis (CRT), which can be used to develop classification for predicting the group membership of commercial banks into two pre-defined groups, namely domestic and foreign banks. The application of the three techniques is illustrated by comparing the classification models obtained by applying them to selected liquidity, cost-revenue, profitability and activity bank ratios data set. As the results reveal that CRT outperform traditional discriminant analysis and logistic regression techniques in terms of bank classification accuracy and thus provide an effective alternative for imple-menting bank classification tasks.
Logistic Regression and Discriminant Analysis, Data Mining, CRT and Bank Performance.

Tam Metin: PDF