We consider the problem of separating two sets of points in an n-dimensional real space with a (hyper)plane that minimizes the sum of L2-norm distances to the plane of points lying on the wrong side of it. We propose an implementation of a nonconvex quadratic programming algorithm based on a branch-and-cut approach to solve this problem. Computational results are reported for several publicly available data sets and larger artificial problems. We also explore the accelerating potential of incorpo- rating heuristic bounds in our exact solution approaches.
Published February 2007 , 16 pages