A. Recall, the decision boundary of a classifier is defined as where estimated class labels are ambiguous. Now consider three training points with feature values (x1,x2):(1,0),(1,0) and (0, sqrt(3)) belonging to class 1 , class 2 and class 3 , respectively. i) For a nearest neighborhood classifier, draw the decision boundary and specify the equation(s) associated with the decision boundary. ii) Is the decision boundary linear or non-linear? iii) Assuming probability of a query point belonging to class 1, class 2 and class 3 , is inversely proportional to Euclidean distance of the query point from training points, what will be probability of query point at (0, sqrt(3)/2) belonging to class 1 , class 2 and class 3 ? B. Consider the following training dataset where x is the predictor variable and y is the target value. Assume predicted value at a query point is equal to 1/Euclidean distance weighted average of 3 nearest neighbors, what will be the estimated output for query point at x=5 ? 4+1+2+3=10 Marks