Improving the Accuracy of the CORER Classifier Using a Hybrid Approach

Rule generator classification algorithms have been successfully used in applications of data mining. The objective of this paper is to improve the accuracy of data classification task. For this purpose, we have merged the CORER classifier (Colonial cOmpetitive Rule-based classifiER) with another rule-based classifier. More specially, the extracted rules of the PART algorithms have been used as the initial rules of CORER. In order to approve the proposed approach capability, two different data sets from UCI machine learning database repository have been applied. To assess the performance of improved CORER, we compared our results with some other well-known classifiers, namely C4.5, CN.2, ID3, PART and CORER which brings about superior results. Our findings lead us to believe that the proposed approach may provide better performance for critical fields which need more precise classification algorithms