1. Using thatstream of data and the same assumptions about handling the smoothed and trending portions of the early periods of the data stream, optimize the forecast by sequentially: 1) finding the alpha (to thenearest increment of 0.05) that produces the lowest Se, then 2) holding that alpha constant, and finding the beta (to the nearest 0.05) which provides the overall lowest Se. Remember, when changing alpha or beta using this model, you must change the entire column,not just one row at a time.
2. What were the final optimal alphas and betas?
3. What is the eventual lowest forecast error using this process?
4. How might you proceed to find a forecast with even lower error?
5. What is the point estimate for the next unknown period? What is the 95% confidence interval estimate?

b C d alpha beta actual S t forecast error squared error W N 0.5 0.5 500 0.5 0.5 512 0.5 0.5 533 515 11.0 0.5 0.5 576 551 23.5 574.5 0.5 0.5 614 594 33.4 627.6 0.5 0.5 656 642 40.5 682.3 -40.3 1622.6 CO 0.5 0.5 642 662 30.4 692.5 5.5 29.8 10 0.5 0.5 698 695 31.8 727.0 -27.0 730.8 11 0.5 0.5 700 714 25.0 738.5 -26.5 703.4 12 0.5 0.5 712 725 18.4 743.6 3086.6 1028.9 forecast error 32.0760037240603 95% confidence interval 680.8 to 969.0