| Methodology adopted | Accuracy (%) | Sensitivity (%) | Specificity (%) | Selected features |
| Optimized LVQ (10x CV) [10] | 96.70 | 91.29 | 92.34 | 2, 3, 6 | Big LVQ (10x CV) [10] | 96.80 | 95.23 | 96.10 | 2, 3, 6 | AIRS (10x CV) [10] | 97.20 | 96.92 | 95.00 | 2, 3, 6, 7 | Supervised fuzzy clustering (10x CV) [11] | 95.57 | 98.23 | 97.36 | 2, 3, 6, 7, 8 | Fuzzy-AIS-knn (10x CV) [12] | 99.14 | 99.56 | 100 | 2, 3, 6, 7, 8 | -score + support vector machine [13] | 99.51 | 99.24 | 98.61 | 2, 3, 6, 7 | Association rule + neural network [14] | 97.4 | 93.12 | 91.26 | 2, 3, 6, 7, 8 | Artificial metaplasticity neural network [15] | 99.26 | 100 | 97.89 | 2, 3, 6, 7, 8 | Mean selection method [16] | 95.99 | 93 | 97 | 2, 3, 6, 7 | Half selection method [16] | 96.71 | 94 | 98 | 2, 3, 6, 7, 8 | Neural network for threshold selection [16] | 97.28 | 94 | 99 | 1, 2, 3, 5, 6, 7, 8 | PSO + ELM | 99.62 | 99.61 | 98.93 | 2, 3, 6, 7, 8 | Proposed SRLPSO + ELM | 99.78 | 100 | 100 | 2, 3, 6, 7 |
|
|