Case-Based Reasoning with feature clustering

Case-based reasoning (CBR) is the process of solving new problems based on the solutions of the similar past problems. Selecting important features to perform case retrieving can improve the efficiency of the large-scale CBR. In this paper, we select features based on attribute clustering. The representative attributes found in the clusters are thus used for indexing and representing cases such that retrieving similar cases based on representative attributes can reduce the execution time. In addition, the clustered attributes also make the CBR framework more flexible than other feature selection methods.