Table 3 Performance comparison between proposed SHOSCA and other optimization algorithms, highlighting its excellence in metrics such as accuracy, precision, recall, and F1 score for MS detection. It evidences SHOSCA’s enhanced efficiency in FS and model optimization, despite a longer CPU time, underscoring its potential as a robust tool in clinical diagnostics.

From: AI-based model for automatic identification of multiple sclerosis based on enhanced sea-horse optimizer and MRI scans

 

Feat

Size

Acc

Prec

Rec

F1

Score

CPU

Time

Avr

fit

Std

fit

Best

fit

Worst

fit

SHO

9.48

85.7882

83.3532

79.7576

81.3509

3.9979

0.1213

0.0149

0.0933

0.1514

HHO

98.92

88.6118

86.2826

84.1212

85.1439

4.0070

0.1257

0.0238

0.0699

0.1556

WOA

76

89.4118

88.0679

84.3636

86.0967

1.7474

0.1270

0.0237

0.0815

0.1544

SCA

46.48

89.9294

89.3919

84.1212

86.6184

3.1534

0.1124

0.0145

0.0825

0.1401

Proposed

SHOSCA

63.76

91.2471

90.7619

86.3030

88.4382

8.0064

0.1099

0.0105

0.0944

0.1295

  1. *Significant are in bold.