Sentence examples for model output metrics from inspiring English sources

This possibility was therefore analysed further below.The sensitivity patterns described above were confirmed by the plot of the correlations (Pearson's R) between all input parameters and model output metrics included in this analysis, shown in Figure  5.

PLSR is a subspace-based regression method based on decomposing the data into a subspace representing the main features of covariance between the regressors (here input parameters) and the response variables (here model output metrics).

As described above, in order to re-parameterise the Niederer-model, we used a combination of inverse metamodelling [ 5, 8], predicting the input parameter values directly from the model output metrics, and iterative zooming into the relevant region of the parameter space based on a look-up approach.

Following simulations with the contraction model, the output metrics described above were calculated from the model outputs generated using the parameter values from the experimental design (step 3).

Biological models typically contain numerous output metrics resulting from large sets of coupled equations, and complex covariance patterns often exist between the outputs.

Figure 4 summarizes the model output Cfs for the four exposure metrics.

The next step of the fitting pipeline is to generate an inverse HC-PLSR metamodel, predicting the Niederer model parameters as functions of the output metrics in Table  2 and Table  3, based on the simulation results.

The model can still be sensitive to variations in these parameters, but it is difficult to predict parameter values from the output metrics for sloppy models.

Hence, most of the parameters from the Land-model could be constrained by the output metrics considered.

For this case study, RAIDAR version 2.0 was coded in Visual Basic for Applications hosted in Excel (Microsoft Corp., Redmond, WA, USA) to quantify the propagation of uncertainty in chemical properties (model input parameters) in the calculated iF, iRA, CU, and CA exposure metrics (i.e., model output).

Based on the parameter-simulated output data for the Niederer-model, a classical metamodel was first constructed to predict the output metrics as functions of the parameters using PLSR.