Table 4 Statistical significance testing of RIAM against baseline models (tenfold CV).
Label | Comparison | RIAM (Mean ± Std) | Baseline (Mean ± Std) | ΔMean | Holm-adjusted p | Cohen’s dz | Cliff’s δ | Effect | |
|---|---|---|---|---|---|---|---|---|---|
s1 | RIAM vs LR | 0.9890 ± 0.0056 | 0.9313 ± 0.0134 | 0.0578 | 0.00293 (**) | 4.74 | 1 | Large | |
s1 | RIAM vs NB | 0.9890 ± 0.0056 | 0.9042 ± 0.0105 | 0.0848 | 0.00195 (**) | 9.008 | 1 | Large | |
s1 | RIAM vs SVM | 0.9890 ± 0.0056 | 0.9198 ± 0.0058 | 0.0693 | 0.00098 (***) | 8.838 | 1 | Large | |
s2 | RIAM vs LR | 0.9710 ± 0.0113 | 0.8388 ± 0.0274 | 0.1322 | 0.00293 (**) | 4.5 | 1 | Large | |
s2 | RIAM vs NB | 0.9710 ± 0.0113 | 0.9320 ± 0.0114 | 0.039 | 0.00195 (**) | 3.479 | 0.98 | Large | |
s2 | RIAM vs SVM | 0.9710 ± 0.0113 | 0.9180 ± 0.0011 | 0.053 | 0.00098 (***) | 4.617 | 1 | Large | |
s3 | RIAM vs LR | 0.9908 ± 0.0044 | 0.9376 ± 0.0108 | 0.0532 | 0.01855 (*) | 3.905 | 1 | Large | |
s3 | RIAM vs NB | 0.9908 ± 0.0044 | 0.8811 ± 0.0107 | 0.1098 | 0.01855 (*) | 9.595 | 1 | Large | |
s3 | RIAM vs SVM | 0.9908 ± 0.0044 | 0.9837 ± 0.0086 | 0.0071 | 0.01855 (*) | 0.789 | 0.46 | Medium | |
