Efficient Surrogate Modeling with T-AHS: A Leave-One-Out Cross-Validation Approach

The model employed in this study, T-AHS, has the capability to eliminate underperforming individual models and select appropriate ones by utilizing leave-one-out cross-validation error. The adaptive weight factors for each surrogate model are determined by considering the baseline model and the estimated mean square error in a Gaussian process prediction. Furthermore, T-AHS exhibits shorter computational time compared to other hybrid models.