CBR-Based Control System of Credit Risk
Wei Ren Chen · Journal of Northeastern University · 2007
Taking the complex credit risk control as a background,a new frame of CBR(case-based reasoning) system is built to improve further the validity of the existing CBR system.Introducing the frame-based case representation,a novel case adaptation method based on fuzzy proximity is presented for case retrieval.A new way combining statistic analysis with programming model is proposed for case adaptation,and several methods are discussed to maintain properly the definite cases.By the system running of instances,both the validity and veracity are verified.Compared with the conventional nearest neighbor approach,the adaptation method based on fuzzy proximity can improve the retrieval precision and the new way for case adaptation will improve the quality of solutions.