The BLMM approach resulted in models with high predictive power and low prediction variance, likely due to the utilisation of the correlation of pH values at co-located sites through time.
The test demonstrates that the more prior knowledge may not only improve the prediction accuracy of 3NN fusion strategy in target predicting, but also reduce its prediction variance.
This algorithm chooses runs that minimize the integral of the prediction variance across the factor space.
The proposed criterion is derived by performing an asymptotic expansion to the conditional prediction variance.
A design criterion related to G-optimality is G-efficiency, defined as {text{G-efficiency}} = 100 times left( {frac{p/n}{text{MPV}}} right), (4 where (p/n) is the average prediction variance (APV) and MPV is the maximum prediction variance.
The results show that uncertain inputs and ANN model parameters produces the most prediction variance, followed by prediction variances from uncertain number of hidden layer nodes and uncertain activation function.
The IV optimal design seeks to minimize the integral of prediction variance across the entire design space.
The dissimilarities between the methods for estimating the prediction variance and non-optimal estimated uncertainties were important limitations for the B-G models despite applying a global correction factor for the prediction variances.
The cross-validation technique is used to find the best surrogate that has the smallest prediction variance.
Moreover, meta-modelling-based uncertainty analysis is explored to quantify the prediction variance, which is important for risk assessment.
