Gene Selection via a New Hybrid Ant Colony Optimization Algorithm for Cancer Classification in High-Dimensional Data

Abstract

The recent advance in the microarray data analysis makes it easy to simultaneously measure the expression levels of several thousand genes. These levels can be used to distinguish cancerous tissues from normal ones. In this work, we are interested in gene expression data dimension reduction for cancer classification, which is a common task in most microarray data analysis studies. This reduction has an essential role in enhancing the accuracy of the classification task and helping biologists accurately predict cancer in the body; this is carried out by selecting a small subset of relevant genes and eliminating the redundant or noisy genes. In this context, we propose a hybrid approach (MWIS-ACO-LS) for the gene selection problem, based on the combination of a new graph-based approach for gene selection (MWIS), in which we seek to minimize the redundancy between genes by considering the correlation between the latter and maximize gene-ranking (Fisher) scores, and a modified ACO coupled with a local search (LS) algorithm using the classifier 1NN for measuring the quality of the candidate subsets. In order to evaluate the proposed method, we tested MWIS-ACO-LS on ten well-replicated microarray datasets of high dimensions varying from 2308 to 12600 genes. The experimental results based on ten high-dimensional microarray classification problems demonstrated the effectiveness of our proposed method.

[Retracted] Taxonomy of Adaptive Neuro‐Fuzzy Inference System in Modern Engineering Sciences

“…Model hybrids ABC [57] aABC [43,58], adaptive ABC (AABC) [59], vortex search [60], cooperative ABC (CABC) [61,62], cooperative micro-ABC (CMABC) [63], interval cooperative multiobjective ABC (ICMOABC) [62], ABC-PSO [64], multiobjective directed bee colony optimization (MODBCO) [65], Scoutless ABC [35], directed ABC [66,67] ACO [68] ACOR [36], heuristic-PS-ACO (HPSACO) [69], hybrid ACO [70], ACO-PSO [71], PS-ACO [72], ACO-SA [73], MWIS-ACO-LS [74], hybrid ACO (HAntCO) [75], min-max ant System (MMAS) [72,76], GA-ACO-SA [77], self-adaptive ant colonygenetic hybrid [78], GA-ACO [79], ACS [80], greedy ACS [81] BA [82] Binary BA [83], hybrid BA with ABC [84], BA-HS [85], adaptive BA [86], adaptive multiswarm BA (AMBA) [87], binary BA [83], differential operator & Levy flights BA [87], directed artificial BA (DABA) [88], double-subpopulation Levy flight BA (DLBA) [89], dynamic virtual BA (DVBA) [90], improved DVBA with probabilistic selection [91], island multipopulational parallel BA (IBA)…”

A Tri-Stage Wrapper-Filter Feature Selection Framework for Disease Classification

“…For the DLBCL dataset, the work mentioned in [ 14 , 26 , 30 ] attained an accuracy of 100% but the number of features selected by the authors were greater than the number of features selected by our proposed method. They utilized 16, 6, and 8 features, respectively, whereas the proposed method selected only four features as the optimal feature set.…”

“…For the prostate cancer dataset, Bir-Jmel et al [ 26 ] achieved an accuracy of 100%, selecting 21 features. However, the proposed method also obtained the same accuracy while utilizing only three features.…”

“…Bir-Jmel et al [ 26 ] implemented a novel two-stage hybrid method named MWIS-ACO-LS. In the first stage, they used a combined approach of a new graph-based technique (named MWIS) and Fisher score, which acted as a filter method.…”

“…The authors of the work [ 26 ] implemented the proposed method MWIS-ACO-LS for the prostate cancer dataset, achieving a highest classification accuracy of 100% with 21 selected features. The work performed by Sun et al [ 18 ] implemented the novel PDJE-AR method to the prostate cancer dataset, achieving an average accuracy of 91.2% with four features using C4.5 as classification algorithm.…”