The Development of a Credit Scoring System Based on Case-Based Reasoning with Genetic Algorithm Applied

摘要

This research has developed a credit-scoring system based on case-based reasoning (CBR) with the genetic algorithm (GA) applied in the k-nearest-neighbor (k-NN) technique for case retrieving. This system can search the most similar k cases in terms of certain characteristics to support credit rating for a new applicant. Performance of the system has been presented and examined under human selected factors. In addition to credit rating, this research has applied the same techniques to design a fitness function and develop a prototype to support loan amount rating.