This paper deals with the problem of multivariate affine regression in the
presence of outliers in the data. The method discussed is based on weighted
orthogonal least squares. The weights associated with the data satisfy a s
uitable optimality criterion and are computed by a two-step algorithm requi
ring a RANSAC step and a gradient-based optimization step. Issues related t
o the breakdown point of the method are discussed, and examples of applicat
ion on various real multidimensional data sets are reported in the paper.