(d) (20 points) Suppose we only have four training examples in two dimensions as shown in Fig. The positive samples at 2:1 = (0, 0), 1:2 = [2, 2) and negative samples at 2:3 = (h, 1) and 1:4 = (0, 3) Orl- 0.5 1 1.5 2 2.5 3 i. (10 points) For what range of parameter h > 0, the training points are still linearly separable? ii. (10 points) Does the orientation of the maximum margin decision boundary change as it changes, when the points are separable