In judging the fit of the estimated response function in Example 7.5, you could use MAE (mean absolute error) instead of RMSE. MAE is the average of the absolute prediction errors.

a. When you run Solver with MAE as your objective, do you get approximately the same estimated response function as with RMSE?

b. Repeat part a, but do it with the outliers in parts a and b of the previous problem. Report your results in a brief memo.