Microprocessor manufacturer has kept track of demand for microprocessors for last two years (24 months). Data is provided to you for microprocessor demand. Manager has asked use forecasting methods of simple moving average (2, 3 and 4 period MA) and simple exponential smoothing (with alpha 0.3). Use MAD, MSE, and MAPE to evaluate the accuracy in prediction. Not knowing whether 3 period is the best k, and 0.3 is the best alpha, you have decided to do some trial-and-error analysis to try out different values of k and alpha. Considering that using MAD, it may be easier to explain the differences, you have decided to use MAD to compare your choices. Provide your finding on the best number of moving average periods and best smoothing constant for exponential smoothing. Additionally, you have decided to provide visual comparison. On a chart, plot the actual values to examine whether other method may be suitable. In addition, add the data on new chip introduction to the plot (right click in the plot select data add series use secondary axis to plot it on RHS Y axis). What is your observation?

New Chip New Chip Month Introduced Demand Month Introduced Demand 1 No 4243 13 No 8556 2 No 3745 14 No 11371 3 No 6935 15 Yes 17296 4 No 8920 16 No 9919 5 Yes 12546 17 No 14409 6 No 8938 18 Yes 22780 7 No 6128 19 No 17529 8 No 10196 20 No 16936 9 Yes 15502 21 No 17708 10 No 8396 22 No 18277 11 No 10048 23 No 19714 12 No 8265 24 No 20331 Using Time Trend and Multiple regression models, give the prediction model and provide forecasts. Identify (quantitatively) the model that has a better forecast accuracy. What is your conclusion