Machine learning also has intimate ties to optimization: Quite a few learning challenges are formulated as minimization of some reduction functionality on the training list of examples. Reduction functions express the discrepancy between the predictions from the design being experienced and the actual dilemma instances (for instance, in classificat