Fuzzy-neuro intelligent systems approach to unstructured problems

Abstract

Significant change has been observed in problem-solving perspective in the fields of management science, that is, a shift away from general purpose algorithmic approaches to the use of problem domain-specific knowledge. In the fields of expert systems, this new perspective has been used most extensively. And in the areas of management planning, simulation has become one of the most widely used techniques because the aim of business simulation is to help the manager analyze complex problems and make decisions in a more scientific and objective way. Therefore, to the extend of this views, this study deals with decision support mechanism using the knowledge-based simulation technique in the strategic planning field, one of highly ill-structured problems, and a fuzzy logic-driven multiple knowledge integration framework for dealing with uncertainty and/or fuzziness. In this study, we analyze three issues. The first is an issue of multiple knowledge integration (MKI) for improving the performance of expert systems. To maintain high performance in ill-structured situation, expert systems should depend on multiple sources of knowledge rather than a single kind of knowledge. For this reason, Our major concern is to develop a fuzzy logic-driven framework for MKI, and apply it to stock market stage prediction (SMSP) problem which is one of ill-structured decision problems. Types of knowledge being considered herein are threefold: machine knowledge, expert knowledge, and user knowledge. The target problem domain of this study is one week-ahead stock market stage prediction: Bull, Edged-up, Edged-down, and Bear. The proposed fuzzy logic-driven framework for MKI plays a role of integrating user knowledge, expert knowledge, and machine knowledge. Extensive experiments with real data proved that the proposed fuzzy logic-driven framework for multiple knowledge integration can contribute significantly to improving the performance of expert systems. The second issue is a stra...