2. (a) Define ensemble method. Give one example from the class of an ensemble model that gives improved performance over a base model. (b) You are asked to apply the boosting algorithm on a classification problem shown in Fig. 1 using decision stump(s). (Hint: Decision stump classifier chooses a constant value c and classifies all points where x> c as one class and other points where rc as the other class.) i. What is the initial weight that is assigned to each data point? 0 y=-1 1 + 2 y=+1 + 3 4 y=-1 5 6 Figure 1: Training data for boosting algorithm. ii. Draw and show the decision boundary using Fig. 1 for the first decision stump that indicates the positive and negative side of the decision boundary. iii. Considering the decision boundary of the above, mark the data point whose weight increases in the boosting process. iv. What is the weight that is assigned to each data point after the first iteration of boosting algorithm? (c) List the benefits of using "weak" learners in boosting methods.