For the data given below, squares, triangles, and open circles are three different classes of data in the training set and the diamond () and star () are test points. We denote the total number of training points as N and consider K-nearest-neighbor (KNN) classifier with L2 distance. Use the figure below to answer questions 1.11.3. 1.1 What is the prediction for the test point star when K=4 ? Explain why ( 2 points) 1.2 What is the diamond classified as for K=N ? Explain why Note that the test point star is not included in K=N since it is not a training point ( 2 points) 1.3 Suppose one performs leave-one-out validation (that is, N-fold cross validation) to choose the best hyper-parameter K. List the coordinate ( x,y) of triangles that are correctly classified (as a validation point) in this process for the run with K=1. ( 2 points) 1.4 Is KNN considered a parametric or non-parametric method? Explain why ( 2 points)