An Efficient Method for Associative Classification using Jaccard Measure
Dharini B1, Jayanthi M2, Joyce A3
1Dharini B, Assistant Professor, Department of Computer Science and Enginnering, PSNA College of Engineering and Technology, Dindigul (Tamil Nadu), India.
2Jayanthi M, Assistant Professor, Department of Computer Science and Enginnering, PSNA College of Engineering and Technology, Dindigul (Tamil Nadu), India.
3Joyce A, Assistant Professor, Department of Computer Science and Enginnering, PSNA College of Engineering and Technology, Dindigul (Tamil Nadu), India.
Manuscript received on 19 October 2019 | Revised Manuscript received on 25 October 2019 | Manuscript Published on 02 November 2019 | PP: 3448-3453 | Volume-8 Issue-2S11 September 2019 | Retrieval Number: B15800982S1119/2019©BEIESP | DOI: 10.35940/ijrte.B1580.0982S1119
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Classification is a data mining technique that categorizes the items in a database to target classes. The aim of classification is to accurately find the target class for each instance of the data. Associative classification is a classification method that uses Class Association Rules for classification. Associative classification is found to be often more accurate than some traditional classification methods. The major disadvantage of associative classification is the generation of redundant and weak class association rules. Weak class association rules results in increase in size and decrease in accuracy of the classifier. This paper proposes an efficient approach to build a compact and accurate classifier by using interestingness measures for pruning rules. Interestingness measures play a vital role in reducing the size and increasing the accuracy of classifier by pruning redundant or weak rules. Rules which are strong are retained and these rules are further used to build the classifier. The source of the data used in this paper is University of California Irvine Machine Learning Repository. The approach proposed in this paper is effective and the results show that the approach can produce a highly compact and accurate classifier.
Keywords: Associative Classification, Class Association Rules, Accuracy, Interestingness Measure, Classifier.
Scope of the Article: Classification