*3.2. Numerical Fitting of Porcine Carotid Tissue*

For the mean experimental curves (circumferential and longitudinal) of the distal samples, results are shown in Figure 7. It is worth noting that, in general, it seems not to be a good idea to just compute the mean values of the GENERIC models for each of the neighboring experimental curves. Particularly in the circumferential direction, the deviation from the reference solution is noteworthy. As in the previous section, results provided by Kriging interpolation outperform this approach. Particularly, local Kriging is found to provide the highest accuracy. The predicted behavior is almost indistinguishable from the mean experimental results.

(**b**) Distal-longitudinal samples

**Figure 7.** Comparison of (**a**) distal-circumferential and (**b**) distal-longitudinal models predicted by mean GENERIC values, or by Kriging interpolation of those samples neighboring the reference solution.

With the weights just computed for the distal samples, we constructed a new GENERIC model for the proximal results. Its predictions are shown in Figure 8. Once again, by just computing the mean of the GENERIC terms for the neighboring curves does not seem to produce good results. However, with the Kriging weights computed for the distal samples, results for the proximal samples are equally good. This demonstrates the robustness of the proposed approach.

(**b**) Proximal-longitudinal samples

**Figure 8.** Comparison of (**a**) proximal-circumferential and (**b**) proximal-longitudinal models predicted by mean GENERIC values, or by Kriging interpolation of those samples neighboring the reference solution.

It is worth noting that the computational cost of this procedure is by no means high: each sample took on average 2.52 seconds to obtain the corresponding GENERIC model.
