T datasets, the minimum quantity of capabilities chosen by B-MFO shows
T datasets, the minimum quantity of features selected by B-MFO shows that B-MFO could steer clear of the local optima trapping and acquire the optimum option. Figure four presents the typical quantity of chosen capabilities in large datasets: PenglungEW, Parkinson, Colon, and Leukemia. These results indicate the significant effect of transfer functions on algorithms’ behavior within the position updating of search agents and acquiring the optimum resolution within the feature selection difficulty. Among the three categories of transfer functions employed by B-MFO, the U-shaped transfer functions outperform the V-shaped and Compound 48/80 web S-shaped when it comes to maximizing the classification accuracy and minimizing the amount of selected options, specially for big datasets.Computer systems 2021, ten,11 Tianeptine sodium salt Formula ofTable 3. The accuracy and chosen features’ number gained by winner versions of B-MFO and comparative algorithms. Datasets (Winner) Pima (B-MFO-S1) Metrics Avg accuracy Std accuracy Avg no. options Avg accuracy Lymphography (B-MFO-V3) Std accuracy Avg no. options Avg accuracy Breast-WDBC (B-MFO-U3) Std accuracy Avg no. functions Avg accuracy PenglungEW (B-MFO-U2) Std accuracy Avg no. options Avg accuracy Parkinson (B-MFO-V2) Std accuracy Avg no. attributes Avg accuracy Colon (B-MFO-U2) Std accuracy Avg no. features Avg accuracy Leukemia (B-MFO-U2) Std accuracy Avg no. functions BPSO 0.7922 0.0033 4.7333 0.9163 0.0099 eight.9333 0.9710 0.0021 12.8333 0.9626 0.0040 161.0667 0.7952 0.0243 376.4333 0.9625 0.0056 999.9333 0.9988 0.0013 3542.0670 bGWO 0.7726 0.0063 7.6000 0.8694 0.0108 16.9667 0.9626 0.0028 27.6000 0.9541 0.0044 322.6667 0.7736 0.0036 741.2333 0.9526 0.0048 1948.8667 0.9901 0.0021 6746.9670 BDA 0.7849 0.0119 3.2667 0.9041 0.0182 five.5333 0.9666 0.0078 two.4000 0.9507 0.0126 83.5667 0.7643 0.0056 192.7333 0.9296 0.0207 618.4333 0.9703 0.0167 2283.7330 BSSA 0.7798 0.0079 4.7667 0.8882 0.8882 9.1000 0.9655 0.0030 13.8000 0.9567 0.0058 199.5000 0.7793 0.0126 332.7667 0.9535 0.0051 1152.2000 0.9954 0.0023 3435.2330 B-MFO 0.7902 0.0046 five.2667 0.9095 0.0089 5.3667 0.9719 0.0020 3.2333 0.9692 0.0063 81.5333 0.8603 0.0094 79.1000 0.9694 0.0059 350.7667 0.9998 0.0005 669.Table four. The comparison final results amongst winner versions of B-MFO and comparative algorithms on fitness. Datasets (Winner) Pima (B-MFO-S1) Lymphography (B-MFO-V3) Breast-WDBC (B-MFO-U3) PenglungEW (B-MFO-U2) Parkinson (B-MFO-V2) Colon (B-MFO-U2) Leukemia (B-MFO-U2) Metrics Avg fitness Std fitness Avg fitness Std fitness Avg fitness Std fitness Avg fitness Std fitness Avg fitness Std fitness Avg fitness Std fitness Avg fitness Std fitness BPSO 0.2117 0.0034 0.0878 0.0095 0.0330 0.0019 0.0420 0.0040 0.2078 0.0241 0.0421 0.0055 0.0062 0.0013 bGWO 0.2347 0.0068 0.1387 0.0110 0.0462 0.0027 0.0554 0.0043 0.2340 0.0035 0.0567 0.0048 0.0192 0.0022 BDA 0.2456 0.0052 0.1503 0.0189 0.0571 0.0111 0.8845 0.1006 2.1607 0.2104 6.2540 0.5740 22.8667 two.6745 BSSA 0.2240 0.0076 0.1157 0.0106 0.0387 0.0033 0.0490 0.0059 0.2229 0.0135 0.0518 0.0051 0.0094 0.0023 B-MFO 0.2143 0.0046 0.0925 0.0084 0.0289 0.0021 0.0330 0.0061 0.1393 0.0095 0.0321 0.0056 0.0011 0.Computer systems 2021, 10,12 ofTable five. The comparison benefits involving winner versions of B-MFO and comparative algorithms on specificity and sensitivity.Datasets Metrics (Winner) Computer systems 2021, 10, x FOR PEER Assessment Avg specificity PenglungEW Computer systems 2021, 10, x FOR PEER Assessment (B-MFO-U2) Avg sensitivity Parkinson Parkinson (B-MFO-V2) BPSO 0.9975 0.9722 bGWO 1.0000 0.9444 BDA 0.9940 0.9333 BSSA 0.9980 0.