In this paper we consider the problem of the best approximation of a continuous multivariate function in two spaces L1 and L2 by optimization.