A multi-objective evolutionary algorithm-based ensemble optimizer for feature selection and classification with neural network models

• Published in: Neurocomputing (Amsterdam) Vol. 125; pp. 217 - 228
• Main Authors: Tan, Choo Jun, Lim, Chee Peng, Cheah, Yu–N
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 11.02.2014
• Subjects: Algorithms; Classification; Ensemble models; Evolutionary; Evolutionary algorithm; Evolutionary algorithms; Feature selection; Human motion; Multi-objective optimization; Neural network classifiers; Neural networks; Performance enhancement; Tasks; Ensemble models; Feature selection; Multi-objective optimization; Neural network classifiers; Evolutionary algorithm
• ISSN: 0925-2312; 1872-8286
• DOI: 10.1016/j.neucom.2012.12.057

In this paper, we propose a new multi-objective evolutionary algorithm-based ensemble optimizer coupled with neural network models for undertaking feature selection and classification problems. Specifically, the Modified micro Genetic Algorithm (MmGA) is used to form the ensemble optimizer. The aim of the MmGA-based ensemble optimizer is two-fold, i.e. to select a small number of input features for classification and to improve the classification performances of neural network models. To evaluate the effectiveness of the proposed system, a number of benchmark problems are first used, and the results are compared with those from other methods. The applicability of the proposed system to a human motion detection and classification task is then evaluated. The outcome positively demonstrates that the proposed MmGA-based ensemble optimizer is able to improve the classification performances of neural network models with a smaller number of input features.