Clustering based deletion policy for case-base maintenance

Case-base maintenance (CBM) is becoming more important with the increased use of case-based reasoning (CBR) systems especially in machine learning. Large scale CBR systems are becoming more ubiquitous, with huge sizes of case libraries consisting of thousands to millions of cases. Large case-bases raise the concern about the utility problem for case retrieval and emphasize on the need of controlling case-base growth through certain policies. Various case-base deletion and addition strategies have been suggested which claim to preserve case-base competence. In this paper, we present a clustering based deletion strategy for case-base maintenance which exploits k-means clustering algorithm. The results presented in this paper reveal that the proposed policy performs better than the existing benchmark deletion policy and ensures better competence.