1. Suppose we observe a vector y e R" of observations from the model, y - XB* +E, where X n. Show that there exist a vector v / 0 such that Xv - 0. Argue based on part b, that there are infinitely many linear regression estimates. Furthermore argue that there is a variable i e {1, .."} such that the regression coefficient of variable i can have different signs, depending on which estimate we choose. Comment on this