5. (20 points) After random sampling from an underlying distribution, you obtain a small square uniformly filled with 1000 CLASS 1 examples and a large square uniformly filled with 1000 CLASS 2 examples. a) What is the largest classification accuracy you can obtain by any machine learning method? b) What would be the accuracy of 1-nearest-neghbor classifier? c) What is the best achievable accuracy by a linear classifier? Why