Skip to main content
. 2018 Aug 17;12:327. doi: 10.3389/fnhum.2018.00327

Figure 2.

Figure 2

Ten-fold cross-validated prediction of frustrating and non-frustrating drives from action unit (AU) activations using multivariate logistic ridge regression analysis for an example participant (P11). The data of this participant allowed the classification with the highest accuracy.