Title: An effective feature selection approach driven genetic algorithm wrapped Bayes naïve

Authors: Sidahmed Mokeddem; Baghdad Atmani; Mostefa Mokaddem

Addresses: Department of Computer Science, Computer Science Laboratory of Oran (LIO), University of Oran, BP 1524 El M'Naouer 31000 Oran, Algeria ' Department of Computer Science, Computer Science Laboratory of Oran (LIO), University of Oran, BP 1524 El M'Naouer 31000 Oran, Algeria ' Department of Computer Science, Computer Science Laboratory of Oran (LIO), University of Oran, BP 1524 El M'Naouer 31000 Oran, Algeria

Abstract: In this paper, an advanced novel feature selection (FS) algorithm is presented, the hybrid genetic algorithm (GA) with Bayes naïve (BN), which selects the most relevant optimum feature subset to increase the classification accuracy performance and computational timesaving. Based on GA, the proposed algorithm uses the highest genetic operator's values that involve a consistent classification accuracy. The performance of the algorithm is evaluated using 34 UCI/KEEL datasets from different domains. The classification accuracy is compared with recent literature works. On the other hand, the algorithm is also compared with an implemented sequential forward selection (SFS) technique coupled with different ML methods. The experimental results show the effectiveness of the algorithm.

Keywords: naive Bayes; best first search; BFS; C4.5; feature selection; genetic algorithms; machine learning; multi-layer perceptron; MLP; sequential forward search; support vector machines; SVM; classification accuracy.

DOI: 10.1504/IJDATS.2016.079056

International Journal of Data Analysis Techniques and Strategies, 2016 Vol.8 No.3, pp.220 - 243

Received: 11 Feb 2014
Accepted: 02 Sep 2014
Published online: 12 Sep 2016 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article