3 (3 points) Suppose we are fitting a straight line a+bx to some data (x_i, y_i), i = 1,2,3,4,5. We find the least squares best fitting solution f(x) = a+bx. Write x=(x_1,..,x_5), y=(y_1,..y_5) and A for the matrix whose first column is all 1s and the second column is x. What does "least squares best fit" correspond to? Choose all that apply. Question 3 options: We have minimized the norm of y We have minimized the distance between y and any vector of the form Ar where r is in R^2. We have minimized the sum of the squares of (ax_i - by_i) over all i. We have minimized the sum of the squares of (y_i - f(x_i)) from among all choices of a,b. We have minimized the projection matrix onto the column space of A. We have minimized the square of the norm of x-y We have minimized the distance between y and Col(A) We have minimized for each i the distance from (x_i,y_i) to any straight line of the form a+bx