(a) You are given a data set S = (x1,y1),...,(xt,yt) where each xi and yi are real numbers. You perform simple linear regression to obtain the line 0 + 1x. Now re-scale the y is so that y i = yi for some real number . Perform simple linear regression again. How do the coefficients 0, 1 change for the new line, quantitatively? You may reason by drawing a picture or using formulas for these coefficients from class. (b) What happens if the xi's are scaled by ? (c) For each of the following scenarios, state whether or not we can effectively use linear regression, and give a short reason. (i) We have training data (x,y) (where xR2,yR) satisfying y= x1 + x2, and we want to learn the model parameters ,. (That is, we have training data of the above form for various different x.) (ii) We have training data (x,y) (where xR2,yR) satisfying y= x2 1 + x3 2, and we want to learn the model parameters and . (iii) We have training data (x,y) (where x R2,y R) satisfying y = 2x 1 , and we want to learn the model parameters ,