Question 4: a. Consider the multiple linear regression model Yr = Bo + BiX, +e. It is known that var(e) o Ti. Find a transformation such that the transformed model has constant variance. Clearly defined all the symbols in the transformed model in terms of the symbols given in the original model. b. Explain the what assumptions will be affected by transforming the response variable. c. Explain the what assumptions will be affected by transforming the explanatory vari- ables. d. Assume the residual diagnostics suggested that the model has mean 0, constant vari- ance and normally distributed error. The only problem is that the plot of residual versus the predicted response has a cyclic pattern. Then how would you deal with this problem? e. Assume the normal probability plot of the residuals does not look like a straight line. Then how would you fix this