Support Vector Machines with the Hard-Margin Loss: Optimal Training via Combinatorial Benders' Cuts

Open in new window