Machine learning approach for predicting post-intubation hemodynamic instability (PIHI) index values: towards enhanced perioperative anesthesia quality and safety

Table 4 Performance of the machine learning models on the testing datasets

Models	MAE (CI)	RMSE (CI)	MAPE (CI)	R² (CI)
MLR	0.2042 (0.2042–0.2042)	0.2414 (0.2414–0.2414)	0.7451 (0.7446–0.7456)	0.1151 (0.1150–0.1151)
SVR	0.0960 (0.0958–0.0963)	0.1068 (0.1066–0.1070)	0.3289 (0.3279–0.3299)	0.8067 (0.8260–0.8274)
ETR	0.0512 (0.0511–0.0513)	0.0792 (0.0790–0.0794)	0.2086 (0.2077–0.2095)	0.9047 (0.9043–0.9052)
MLP	0.1212 (0.1180–0.1243)	0.1546 (0.1522–0.1571)	0.4634 (0.4511–0.4758)	0.6366 (0.6251–0.6482)
XGBoost	0.0590 (0.0588–0.0593)	0.0914 (0.0910–0.0918)	0.2263 (0.2250–0.2276)	0.8731 (0.8720–0.8742)

Performance metrics are presented with 95% confidence intervals.
MLR multiple linear regression, SVR support vector regression, ETR extra tree regression, MLP multilayer perceptron, XGBoost extreme gradient boosting, MAE mean absolute error, RMSE root mean square error, MAPE mean absolute percentage error, R² R-squared index, CI confidence interval.
Bold values represent the best performance, indicating that the ETR model outperformed the other models with higher accuracy and reliability.

ISSN: 1471-2253