For investigating the accuracy of the proposed models the Mean Absolute Error (MAE), the Root Mean Squared Error (RMSE) and the Mean Absolute Percentage Error (MAPE) were adopted.
The external prediction accuracy for binary models was as high as 91 96%; for continuous models the mean coefficient R2 for regression between predicted versus observed values was 0.76 0.79.
The ARMA-GARCH model models the mean equation via an (operatorname{ARMA}(p,q)) model, see Eq. (1), whereas the random noise components, represented by (epsilon_{t}), are modelled with a Generalized Autoregressive Conditional Heteroskedastic (GARCH), model.
For the PSiHB models, the mean interatomic distance between first silicon neighbors is 2.35 Å, the second silicon neighbors are distributed in two peaks at 3.25 and 3.85 Å, and the third neighbors are at a distance of 4.85 Å.
The mean value of optimised k for models using random forest was 18 (range 8 48) while for k-MSN models, the mean value of k was 35 (range 9 58).
In these models, the mean variation of the nonlinear soil properties follows a normal distribution; the standard deviation of the normalized shear modulus ((sigma_{text{NG}} )) and the standard deviation of the damping ratio ((sigma_{text{D}} )) are predicted by Eqs.
