In this problem, you perform binary classification given a dataset that consists of m training examples {(x^(i),y^(i))}. We formulate the following soft-margin primal by introducing the slack variable i min(w,b) ||w||2^2 + c sigma(i = 0, m) i Subject to y^(i) (wTx^(i) + b) >=2 - i, i >=0 (for every 1 <= i <= m) Instead of solving the primal, we solve the corresponding dual, thereby learning the optimal dual a and then recovering optimal primal w*,b*. and i*, for every i as a reusit of training. The following figure visualizes the learned decision boundary (blue solid line) and its margin borders (blue dotted lines). Among our m training examples, the selected 12 examples are drawn. - Positive example: Green + - Negative example: Red o