FIG. 8.

Feature selection 1: $u_t=f_1^u(u,u_{xx},v)$ and $v_t=f_1^v(u,v,v_{xx})$. These selected variables are the same as those that appear in the right-hand side of the explicitly known FHN PDE. Regression results of the two methods for time derivatives: Gaussian processes (GPs) and neural networks (NNs).