Title: A novel ensemble classifier by combining sampling and genetic algorithm to combat multiclass imbalanced problems
Authors: Archana Purwar; Sandeep Kumar Singh
Addresses: Department of Computer Science and Information Technology, JIIT Noida, India ' Department of Computer Science and Information Technology, JIIT Noida, India
Abstract: To handle datasets with imbalanced classes is an exigent problem in the area of machine learning and data mining. Though a lot of work has been done by many researchers in the literature for two-class imbalanced problems, the multiclass problems still need to be explored. In this paper, we propose sampling and genetic algorithm based ensemble classifier (SA-GABEC) to handle imbalanced classes. SA-GABEC tries to find the best subset of classifiers for a given sample that is precise in predictions and can create an acceptable diversity in features subspace. These subsets of classifiers are fused together to give better predictions as compared to a single classifier. Moreover, this paper also proposes modified SA-GABEC which performs the feature selection before applying sampling and outperforms SA-GABEC. The performance of the proposed classifiers is evaluated and compared with GAB-EPA, Adaboost and bagging using minority class recall and extended G-mean.
Keywords: feature extraction; diversity; genetic algorithm; ensemble learning; multiclass imbalanced problems.
International Journal of Data Analysis Techniques and Strategies, 2020 Vol.12 No.1, pp.30 - 42
Received: 27 Mar 2017
Accepted: 26 Dec 2017
Published online: 10 Feb 2020 *