Hybrid fuzzy-genetic algorithm to automated discovery of prediction rules

Article type: Research Article

Authors: Fadel, Ibrahim A.^{a; *} | Alsanabani, Hussein^a | Öz, Cemil^a | Kamal, Tariq^{b; c} | İskefiyeli, Murat^a | Abdien, Fawzia^b

Affiliations: [a] Department of Computer Engineering, Sakarya University, Sakarya, Turkey | [b] Department of Electrical and Electronics Engineering, Sakarya University, Sakarya, Turkey | [c] Research Group in Electrical Technologies for Sustainable and Renewable Energy (PAIDI-TEP-023), Department of Electrical Engineering, University of Cadiz, Higher Polytechnic School of Algeciras, Algeciras (Cadiz)

Correspondence: [*] Corresponding author. Ibrahim A. Fadel, Department of Computer Engineering, Sakarya University, Esentepe Campus 54187 Serdivan, Sakarya, Turkey. E-mail: ibrahim.fadel@ogr.sakarya.edu.tr.

Abstract: Genetic algorithm is one of data mining classification techniques and it has been applied successfully in a wide range of applications. However, the performance of Genetic algorithm fluctuates significantly. This research work combines Genetic algorithm with fuzzy logic to adapt dynamically crossover and mutation parameters of Genetic algorithm. Two different datasets are taken during the experiment. Several experiments have been performed to prove the effectiveness of the proposed algorithm. Results show that the rules generated from a proposed algorithm are significantly better with high fitness and more efficient as compared to a normal Genetic algorithm.

Keywords: Data mining, hybrid fuzzy genetic algorithm, prediction rules, growth rate

DOI: 10.3233/JIFS-182729

Journal: Journal of Intelligent & Fuzzy Systems, vol. 40, no. 1, pp. 43-52, 2021