Figure 10.
The image displays five line graphs labelled, a, through, e, each illustrating the relationship between maximum depth represented on the horizontal axis and normalised root mean square error percentage, N R M S E, presented on the vertical axis. Each graph features two lines, a solid blue line indicating the training data and a dashed red line representing the validation data. The graphs exhibit a downward trend in N R M S E as maximum depth increases, with variations evident across the different graphs. Additionally, some graphs include vertical lines at specific points of maximum depth, indicating important transitions in the data. The horizontal range for maximum depth is from 0 to 15, while the vertical range for N R M S E is from 0 to 80 percent. The axes are clearly labelled for easy reference, and there is a consistent visual format across all graphs.Optimisation of max_depth for total shrinkage prediction: (a) DT; (b) RF; (c) ET; (d) GB; (e) XGBoost