A Competence-based Deletion Model for the Improvement of Case-based Maintenance in Case-based Reasoning
Case-based reasoning is a methodology of reasoning used to fix new problems in previous experiences. It applies artificial intelligence, machine learning, systems based on knowledge and other associated fields due to both its extensive usage by humans and its appeal as a methodology for building intelligent systems. This paper focuses on two core problems - increasing case base size and decreasing the competency of the case base. These problems arise due to the repetition of the cycles of case-based reasoning. In order to solve these problems, case-based maintenance methods are crucial for case-based reasoning. In this paper, a useful model for case-based maintenance is proposed in order to deliver better outcomes in comparison with random deletion, utility deletion, and footprint utility deletion. In order to study the efficiency of each method, seven datasets from the machine learning repository are applied to each algorithm. The results of the proposed model show that not only is storage size lower but also competence is higher in comparison with other methods after reduction. Moreover, the reduction rate is higher, and problem solving related to performance is significantly better than when other methods are used.