Table 2 Models used and their performances
Authors (year) | Model | Accuracy | AUC | Sample size | Validation |
|---|---|---|---|---|---|
A. Kabeshova (2015) [21] | NEAT | 88% | 0.7 | 3289 | Holdout |
T. Deschamps (2016) [26] | Decision tree | 82% | N/A | 426 | Holdout |
A. Kabeshova (2016) [20] | ANFIS | 87% | 0.7 | 3525 | Cross-validation |
M. L. Homer (2017) [24] | Lasso logistic regression | N/A | 0.71 | 120,881 | Holdout |
J. Razmara (2018) [19] | ANN (psychological and public factors) | 91% | N/A | 200 | Holdout |
B. W. Patterson (2019) [29] | Random forest | N/A | 0.78 | 9687 | Holdout |
J. A. Womack (2020) [22] | Logistic regression | N/A | 0.76 | 275,940 | Holdout |
M. Suzuki (2020) [28] | CNN “triple factor” (mini-mental state + normalized knee extension strength + FIM locomotion) | 65% | N/A | 42 | Cross-validation |
H. Nakatani (2020) [23] | MCMC | N/A | 0.83 | 743 | Holdout |
K. Makino (2021) [27] | Decision tree | 65% | 0.7 | 2520 | Cross-validation |
C. H. Liu (2021) [25] | Bagging + SVM | 71% | 0.72 | 108,940 | Cross-validation |
W. M. Chu (2022) [31] | Random forest | 73% | 0.69 | 1101 | Cross-validation |
Y. Suzuki (2022) [32] | Logistic regression | 77% | 0.75 | 226 | Cross-validation |
R. Thapa (2022) [30] | eXtreme Gradient Boosting | N/A | 0.85 | 2785 | Holdout |
K. N. K. Fong (2023) [33] | Decision tree | 77% | N/A | 1151 | Cross-validation |
J. H. Choi (2023) [35] | XGBoost | N/A | 0.85 | 1090 | Cross-validation |
V. Sharma (2023) [36] | CatBoost | N/A | 0.7 | 224,445 | Holdout |
O. A. Donoghue (2023) [34] | Conditional inference forest | 67% | 0.69 | 4706 | Cross-validation |
E. Soylemez (2024) [37] | Support vector machine (SVM) and Naïve Bayes | 100% | 1 | 120 | Holdout |