To examine the predictive performance of the model the differences between the predicted and observed EQ-5D scores at the individual level were examined by computing the mean squared error (MSE) and root MSE (RMSE).
By using a relatively small amount of experimental data (25 experimental data points in the training set), a very accurate prediction of the retention (percentage normalized differences between the predicted and the experimental data less than 0.6%) was obtained.
The mean absolute deviation (MAD) aggregated the information on differences between the predicted and the observed stage distributions for comparison of the different methods.
Table 4 shows the prediction accuracy of different tolerance values, ranging from 0 to 3. If 1, 2, or 3 differences between the predicted contact number and the observed value are tolerated, the accuracy could reach 0.63, 0.83, and 0.93, respectively.
The differences between the predicted and experimental values were less than 5%.
The resulting differences between the predicted and measured values are generally within 10%.
Error estimates are made based on the differences between the predicted and measured local strains.
However, differences between the predicted performance of source zone partial mass reduction and the known true performance were reasonably small.
The mean differences between the predicted 1/3 octave band insertion loss values behind the screen and the corresponding measured results are within 2 dB.
The non-significant (p > 0.05) differences between the predicted and experimental values of the response variables endorsed the adequacy of the suggested fitted models.
We also explored two determinants of benefit capture with a linear stepwise regression, analysis of differences between the predicted recreation potential and actual use.
