In Method 2 for minimization, how is the problem transformed before applying the standard maximization simplex?