A computational model of a viewpoint-forming process in a hierarchical classifier system
Takahiro Yoshimi, Toshiharu Taura · 1999
In an environment where input information to machine learning (ML) systems using production rules has many and the amount is huge enough, the authors aim for an ML systems with effective performance in finding solutions. When finding solutions, all of the properties of the input information arc not always required. Therefore, the authors assume that a certain mechanism which can select specific properties to be focused on will contribute to this purpose. For the realization and discussion of this mechanism, the authors have focused on the Classifier System (CS) which has more advantages than other ML systems. From the authors' point of view, operation processes in the CS arc thought to involve this mechanism. However, the CS also involves such duality that both the optimization processes of rules for solution finding and the abstraction processes of input information arc in a single process, which may lead to problems. In this paper, the authors propose a computational model in which these two processes arc explicitly separated. The key concept of the proposed model is the Viewpoint-Forming Process for the purpose of using rules for selecting properties to be focused on. This is separate from the standard rules for finding solutions. A computer system is developed to evaluate the utility of this model. The results acquired by applying the model to an example problem arc reported here.