A case retrieval algorithm based on correlation ana lysis
Bing Jiang, Jiankang Liu, Xiaoqiang Zeng · 2014
The key of case-based reasoning process (CBR) is th e case retrieval. With the increase of the number o f cases, the efficiency of the case retrieval decreases. In orde r to ensure efficiency and stability of the CBR sys tem, related clustering algorithms are introduced to make effect ive classification and improve the efficiency of th e case retrieval. Results of many clustering algorithms are affected by selection of initial values. For example, differ ent results of classification may be produced with the same case l ibrary. Therefore, it is probable that the most sim ilar cases cannot be retrieved and the optimal number of case categor ies cannot be determined in common clustering algor ithms. A case retrieval algorithm based on correlation analysis i s proposed according to the process of case-based r easoning. In the algorithm, the gray correlation analysis is adopted to classify cases stored in the case library of the CBR system, and the method oft-distribution in mathematical statist ics is adopted for determining the optimal number ocase categories. Then, corresponding algorithms for case classification and retrievals are designed. Finall y, comparison experiments are made to verify the stability and ef fectiveness of the algorithm. Theoretical analysis and experimental results show that with the number of cases in the c ase library, the algorithm has better stability and efficiency compared with classic algorithms. The model of the algorithm has some practical value.