Pattern recognition for evaluator errors in a credit scoring model for technology-based SMEs

Abstract

A credit scoring model for technology-based small and medium enterprises presupposes evaluator objectivity and evaluation consistency; however, there is always some amount of error in any technology evaluation. This can be due in part to the subjective evaluation attributes that comprise part of the credit scoring model. The evaluated values of subjective attributes can vary among evaluators. In this study, we identified the significant characteristics of both evaluator and evaluation teams in terms of evaluation error using a decision tree analysis. Our results can improve the accuracy of a wide range of evaluation procedures for technology financing.