A binary ABC algorithm based on advanced similarity scheme for feature selection

Hançer, Emrah; Xue, Bing; Karaboğa, Derviş; Zhang, Mengjie

doi:10.1016/j.asoc.2015.07.023

Cited by 153 publications

(51 citation statements)

References 63 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Readers are referred to the Discussion Section for more details on parameter settings. We compared the proposed TCbGA algorithm to several state-of-the-art feature selection algorithms, including DEMOFS [34] , MOEA/D [35], MDisABC [36], W-QEISS [37], SB-ELM [38], HPSO-LS [39], MoDE [40], GASNCM [41], GCACO [42], GCNC [43], UFSACO [44], FFW-DGC [45], QIFS [46], FSFWISIW [47], BALO [48], MI-SC [49], VMBACO [50], HDBPSO [51] and bGWO [52]. Table 3 shows the mean and standard deviation of the classification accuracy obtained by applying our algorithm to each dataset 25 times and the performance of other algorithms reported in the literature.…”

Section: Experiments and Resultsmentioning

confidence: 99%

A tribe competition-based genetic algorithm for feature selection in pattern classification

Xia

2017

Applied Soft Computing

View full text Add to dashboard Cite

Abstract-Feature selection has always been a critical step in pattern recognition, in which evolutionary algorithms, such as the genetic algorithm (GA), are most commonly used. However, the individual encoding scheme used in various GAs would either pose a bias on the solution or require a pre-specified number of features, and hence may lead to less accurate results. In this paper, a tribe competition-based genetic algorithm (TCbGA) is proposed for feature selection in pattern classification. The population of individuals is divided into multiple tribes, and the initialization and evolutionary operations are modified to ensure that the number of selected features in each tribe follows a Gaussian distribution. Thus each tribe focuses on exploring a specific part of the solution space. Meanwhile, tribe competition is introduced to the evolution process, which allows the winning tribes, which produce better individuals, to enlarge their sizes, i.e. having more individuals to search their parts of the solution space. This algorithm, therefore, avoids the bias on solutions and requirement of a pre-specified number of features. We have evaluated our algorithm against several state-of-the-art feature selection approaches on 20 benchmark datasets. Our results suggest that the proposed TCbGA algorithm can identify the optimal feature subset more effectively and produce more accurate pattern classification.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

A tribe competition-based genetic algorithm for feature selection in pattern classification

Xia

2017

Applied Soft Computing

View full text Add to dashboard Cite

show abstract

“…Subsequently, Hancer et al (2015) develop a novel binary ABC algorithm based on advanced similarity scheme for feature selection. In this study, the DE operators such as mutation and recombination are employed instead of the ABC neighborhood search mechanism.…”

Section: Hybrid Methods Based On Abc and De Algorithmsmentioning

confidence: 99%

“…The proposed method is compared with well-known feature selection algorithms such as chi-square (CHI), information gain (IG) and correlation feature selection (CFS), and some recent feature selection methods that involve metaheuristics including harmony search and stochastic search algorithms for feature selection (Nekkaa and Boughaci, 2015), a binary ABC with DE operators (Hancer et. al, 2015), a novel system for feature selection based on gravitational search (Xiang et.…”

Section: Introductionmentioning

confidence: 99%

A hybrid approach of differential evolution and artificial bee colony for feature selection

Zorarpacı

Özel

2016

Expert Systems with Applications

316

117

View full text Add to dashboard Cite

“…Due to certain limitations in the microarray dataset such as curse of dimensionality, the modest numbers of irrelevant genes samples create unwanted noise which leads the classification as a challenging task for given sample FU, 2004. In classification and clustering phase, computational complexity and the dimensionality of the gene expression matrix is increased by the irrelevant genes GAN, 2008;GARRO, 2016, GHORAI, 2010HANCER, 2015. As an outcome, it is very important to include those least correlated genes with the more informative genes, which is a feature selection problem in microarray data analysis.…”

Section: Problem Statementmentioning

confidence: 99%

A novel microarray gene selection and classification using intelligent dynamic grey wolf optimization

Uthayan¹

2019

Genetika

View full text Add to dashboard Cite

Uthayan. K.R. (2019): A novel microarray gene selection and classification using intelligent dynamic grey wolf optimization. - Genetika, Vol 51, No.3, Effective diagnosis of cancer in the medical field is very important to specific treatment. Exact prediction of different cancer types will provide a better treatment and minimization of toxicity in patients. Microarray high dimensionality of gene expression dataand large number of genes against small sample size, noise and repetition in datasets are the main issues which lead to poor classification accuracy. The selection of informative genes and to reduce dimensionality, Gene Selection technique is used in Microarray. In this paper, a novel meta-heurists algorithm based on Grey Wolf Optimization (GWO) and Artificial Intelligence (AI) is combined to design a model for cancer classification. This proposed work consists of two stages. First, a filter method such as Laplacian and Fisher score, are applied to extract the significant subset of features for faster classification and then Intelligent Dynamic Grey Wolf Optimization (IDGWO) is employed to identify the relevant genes. GWO is a swarm-based algorithm selected for gene expression data classification problem, because it makes classification easy about training and testing cancer data. The significant differences between filter methods of datasets are found by using several analyses. The proposed method was applied on five benchmark datasets by considering top 100 ranked genes selected by fisher score in Lymphoma and SRBCT that had a 100% performance using the IDGWO classifier.

show abstract

A binary ABC algorithm based on advanced similarity scheme for feature selection

Abstract: Please cite this article in press as: E. Hancer, et al., A modified binary ABC algorithm based on advanced similarity scheme for feature selection, Appl. Soft Comput. J. (2015), http://dx.

Cited by 153 publications

References 63 publications

A tribe competition-based genetic algorithm for feature selection in pattern classification

A tribe competition-based genetic algorithm for feature selection in pattern classification

A hybrid approach of differential evolution and artificial bee colony for feature selection

A novel microarray gene selection and classification using intelligent dynamic grey wolf optimization

Contact Info

Product

Resources

About