Fig. 9
ANN–LMB training diagnostics and performance evaluation. (a) Error histogram displaying prediction residuals across training, validation, and test sets. (b) Mean squared error (MSE) over training epochs, highlighting the best validation performance at epoch 511. (c) Convergence metrics including gradient norm, damping parameter \(\mu\), and validation checks. (d) Regression plots showing the correlation between target and predicted values for training, validation, test, and combined datasets with \(R \approx 1\) in all cases.
