Elastomagnetic Sensor-Based Long-Term Tension Monitoring of Prestressed Bridge Member with Temperature Compensation

Table 3

Cross-validation and test results using full-scale datasets and derived hyperparameters.

Category		Extra trees (ET)		Random forest (RF)		Gradient boosting regression (GBR)
		MAE (ton)	R^2	MAE (ton)	R^2	MAE (ton)	R^2
Sensor 1	Validation	0.0026	0.9921	0.0043	0.9793	0.0094	0.9838
	Test	0.0006	0.9992	0.0028	0.9867	0.0105	0.9570
Sensor 2	Validation	0.0025	0.9844	0.0032	0.9772	0.0085	0.9825
	Test	0.0065	0.9373	0.0089	0.8908	0.0123	0.9371
Tuned hyperparameter		Max_features: 1.0, min_samples_leaf: 1, min_samples_split: 2, n_estimators: 100		Bootstrap: true, max_features: 1.0, min_samples_leaf: 1, min_samples_split: 2, n_estimators: 100		Max_features: 1.0, min_impurity decreases: 1e − 09, min_samples_leaf: 1, min_samples_split: 10, n_estimators: 300