HOMEWORK 2 Following this worksheet (using the tabs at the bottom of this page), you will find several worksheets. Each includes a series of demand data. On each sheet: 1. Use the labeled method to fit a forecast to this data. 2. Plot the actual demand vs. the forecast. The first one is done for you as an example (as you input forecast values, the forecast plot will show up). 3. Calculate the various error measures at the bottom of the sheet. 4. Answer any questions that appear at the bottom of the sheet. Hints: My advice is to put the parameters (alpha, beta, gamma, etc) into individual cells and then just refrence those cells in your formulas. That way you can easily change the values of the parameters and see the impact on the forecast and the error measures. The output of the forecast should be in the same units as the actual demand Also, don't forget to round the output of your forecast to match the input of actual demand ("=round()") Holt-Winter's with m=2 Forecast 2004 Ship- ments 6.7 5.6 6.0 8.0 5.5 5.1 lococo O U U Ulooo 6.9 2005 Date Sept Oct Nov Dec Jan Feb Mar Apr |May Jun Jul Aug Sept Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug U 5.5 5.5 7.6 5.9 6.0 7.9 7.2 7.1 9.1 6.8 6.7 8.5 6.9 7.0 8.6 7.0 7.2. NNN OO O O O CONLOO 2006 GO 1. What is the MPE of the forecast? 2. What is the MAE of the forecast? 3. What is the U-stat of the forecast? Are you surprised that it is less than one? 4. Find the best values for alpha and beta that you can. How did you do it