A Preprocessing Model for Integrating Case-Based Reasoning and Prototype-based Neural Network

Generally, a Case-Based Reasoning (CBR) system is divided into two major modules: a case base and a problem solver. Cases in the case base describe the previous solved problems. This base forms the memory or the library of the CBR system. The problem solver in its turn is composed of two components: a case retriever and a case adapter. Given a new input problem, the case retriever identifies the most appropriate cases (in some similarity sense) in the case base and feeds them to the case adapter. The case adapter then examines the retrieved cases and tries to solve the new problem by adapting these cases (Aamodt and Plaza 1994), (Barletta 1991), (Riesbeck and Schank 1989).