Table 4 Friedman test results for the MS detection method (Hybrid SHOSCA) versus other optimization approaches based on performance metrics like feature size, accuracy, precision, recall, F1 score, CPU time, and average fitness. SHOSCA consistently ranks highest in performance metrics, highlighting its effectiveness in diagnosis despite a longer CPU time. The statistical significance of these outcomes, as shown by low p-values, underscores SHOSCA’s superior performance in enhancing MS detection accuracy.
Proposed SHOSCA | SHO | Hybrid Motor | Hybrid breeding | HHO | WOA | SCA | P-value | |
|---|---|---|---|---|---|---|---|---|
Feat Size | 5.84 | 2.38 | 2.54 | 2.22 | 5.12 | 4.9 | 5 | 1.11E-14 |
Accuracy | 6.06 | 3.1 | 2.68 | 1.8 | 4.46 | 5 | 4.9 | 2.22E-14 |
Precision | 6.2 | 3 | 2.7 | 2.02 | 3.82 | 5.04 | 5.22 | 1.90E-14 |
Recall | 5.64 | 3.36 | 2.7 | 1.7 | 4.78 | 5.08 | 4.74 | 1.87E-13 |
F1 score | 5.92 | 3.16 | 2.54 | 1.94 | 4.66 | 4.92 | 4.86 | 4.93E-13 |
CPU Time | 7 | 4.36 | 3.24 | 5.84 | 4.2 | 1 | 2.36 | 3.03E-26 |
Avr fit | 2.88 | 3.92 | 4.06 | 3.94 | 4.96 | 4.64 | 3.6 | 2.17E-02 |
