TABLE 4.

Hybrid Supervised Gene Selection.

References	Ideology	Gene Selection Algorithm	Classifier	Dataset	Performance Evaluation Metrics
Zare et al., 2019	Addresses the linear independence to find informative features with the help of matrix factorization and SVD.	Matrix Factorization based on SVD	Naïve Bayes, C4.5, and SVM	• Brain • CNS • Colon • DLBCL • GLI • Ovarian • SMK • Breast • Prostrate	• Cross-Validation (5-Fold and DOB-SCV) • Sensitivity • Specificity • Accuracy • G-Mean
Chinnaswamy and Srinivasan, 2016	The correlation coefficient is used as the attribute evaluator and PSO as a search strategy to select the necessary features.	Correlation Coefficient and PSO	ELM, J48, Random Forest, Random Tree, Decision Stump, and Genetic Programming	• SRBCT • Lymphoma • MLL	• Classifier Accuracy
Alshamlan et al., 2015	The Genetic Bee Colony combines the benefits of the Genetic Algorithm and Artificial Bee Colony. The method is evaluated using SVM.	Genetic Bee Colony	SVM	• Colon • Leukemia • Lung • SRBCT • Lymphoma	• Classification Accuracy • LOOCV
Liao et al., 2014	Two-stage feature selection methods involve the Laplacian Score and wrapper approach (SFS and SBS) to select the superior genes. Also, it considers the variance information.	Locality Sensitive Laplacian Score, Sequential Forward Selection and Sequential Backward Selection	SVM	• Acute Lymphoma • Lung Cancer • DLBCL • Prostrate • MLL Leukemia • SRBCT	• Accuracy • Precision • Recall • F-Score • AUROC
Shukla et al., 2018	This hybrid method targets at improving the classification accuracy with a two-stage method. It comprises the EGS (multi-layer and F-Score approach) as the first stage to reduce the noise and redundant features; in the second stage, AGA is used as a wrapper to select the informative genes used SVM and NB as fitness functions.	Multi-Layer Ensemble Gene Selection (EGS) and Adaptive Genetic Algorithm (AGA)	SVM and Naïve Bayes	• Breast • Colon • DLBCL • SBRCT • Lung • Leukemia	• Accuracy • FMeasure • Sensitivity
Sun et al., 2019a	A hybrid gene selection method combining the ReliefF and the Ant Colony Optimization is proposed. It is a filter-wrapper based gene selection.	ReliefF-Ant Colony Optimization	RFACO-GS	• Colon • Leukemia • Lung • Prostrate	• Classification Accuracy