Energy minimization method for scattered data Hermite interpolation (Q2483192)

When interpolating given functional data and possibly derivative values, one often seeks interpolants minimizing energy in some sense. Useful approaches to this problem are for example polynomial spline or radial basis function methods if the dimension is more than one. In this paper, Hermite interpolation is pursued for this case in two dimensions. The data are given at scattered points in a domain, and a triangulation is used to define the solution spline. Convergence is also studied, and there are examples to demonstrate the usefulness of the approach. An algorithm to compute the spline is also provided.