A new memetic approach for the classification rules extraction problem. (1st January 2013)

• Record Type: Journal Article
• Authors: Benkhider, Sadjia; Drias, Habiba
• Is Part Of: International journal of data mining, modelling and management. Volume 5:Number 4(2013)
• Page Start: 318
• Page End: 332
• Publication Date: 2013-01-01
• Subjects: data mining -- evolutionary computation -- extraction of classification rules -- Michigan approach -- memetic approach -- meta-meta hybridisation; Data mining -- Periodicals; Information science -- Periodicals; Databases -- Periodicals; 005.7
• Languages: English
• ISSNs: 1759-1163

Abstract

This paper presents a memetic algorithm applied to the classification rules extraction problem. In our new approach, our aim is to obtain a better results accuracy relatively to that obtained by a standard genetic algorithm (GA). A memetic algorithm is based on a GA which is improved by hybridising a local search approach. We made a hybrid method to compute a model of classification. In the literature, there are many hybridisation forms: in this paper, we have chosen to make our local search algorithm also based on a genetic approach so our hybridisation is purely evolutionary.