Problem 4.1 [18pts] (Impact of initialization on k-means) Consider a one-dimensional dataset of n = 2m+1 points (m > 1) spread on the real line with one point at x = 0, m points at x = a, and m points at x = (a + b) where 0 1) spread on the real line with one point at x = 0, m points at x = a, and m points at x = (a + b) where 0