Please answer both parts of the following question:

In an experiment, we bake silicon wafers at various temperatures 3:;- and then meaSure the weight 9,; of the correSponding wafers. Recall that a regression model attempts to nd parameters ,81 and g which minimizes the total \"least squares\" distance Q between data points ($1,191), . . \"(ram ya) and the tted line. The regression analysis also gives an estimate 63 of the error variance. (a) (b) (6 points] In the regression method studied in class, we t a line by minimizing sum of the Squares of the vertical deviations of the data points to the tted line. A more natural method might be to minimize the total distance from the tted line to the points1 given by: (1] D = E distul'i. Eh): (50 + 31 3)], i=1 1Where the inner term is the distance from the point (scam) to the line dened by g and 231- Explain why we use the sum of squares vertical deviation instead of (1). (6 points} One of our assumptions regarding the regression model is that the error terms a,- are normal with zero mean and a variance of 0'2. Explain why it is reasonable to assume that the error terms have a mean of U (in other words, Why we can make the zero mean assumption Without loss of generality}