Case-based reasoning: An intelligent approach applied for financial crises warning

Abstract

This paper presents case-based reasoning approach for financial crises warning. Unlike problem solving processes of the traditional rule-based reasoning approach, in which users directly obtain the problem's solution by rule series, CBR is a methodology providing adaptive solu-tions by systematic comparison between current situation and the simi-lar cases stored in the case library. Three critical techniques of CBR are studied intensively in this paper, namely: knowledge representation, re-trieval of similar cases and case learning. At the end of the paper, an implementation of the Case-based Financial Crises Warning Prototype System (CFCWPS) is described in summary. © Springer-Verlag 2003.