Question 4 (K-Nearest Neighbour) Consider that the normalised training set below is part of a K-Nearest Neighbour "model": @relation dummy-data-set @attribute inputl numeric @attribute input2 {a,b,c} @attribute output numeric @data 0.5, a, 10 0, c. 15 1. b, 20 Assume that the input attributes are the first and the second attributes of the training set. What output value would a 3-nearest neighbour (K-3) algorithm predict for an instance with normalised input attributes (0.6, a)? Explain the results. (20 marks)