Calculate Forecasting Results With Weighted Moving Averages And Exponential Smoothing
Question
A teacher wanted her students; X, Y, and Z, to forecast the demand for January 2016 and give two measures of performance of their forecasts. More specifically, she wanted each student to distinctly select one of three forecasting methods (two students cannot select the same method); simple moving averages of order k, exponential smoothing, or weighted moving averages, and the required measures are Cumulative Error and MAD using the data below:
Month Demand Month Demand Month Demand
|
Jul-14 |
56 |
|
Jan-15 |
|
72 |
|
Jul-15 |
|
76 |
|
Aug-14 |
61 |
|
Feb-15 |
|
75 |
|
Aug-15 |
|
75 |
|
Sep-14 |
55 |
|
Mar-15 |
|
70 |
|
Sep-15 |
|
70 |
|
Oct-14 |
70 |
|
Apr-15 |
|
69 |
|
Oct-15 |
|
74 |
|
Nov-14 |
66 |
|
May-15 |
|
73 |
|
Nov-15 |
|
76 |
|
Dec-14 |
65 |
|
Jun-15 |
|
75 |
|
Dec-15 |
|
80 |
The three students submitted the following results:
|
|
X |
Y |
Z |
|
Dec-16 Forecast |
76.07 |
81.00 |
75.00 |
|
Cumulative Error |
35.50 |
23.00 |
35.75 |
|
MAD |
3.88 |
4.00 |
3.45 |
The teacher gave one of them full mark, another one got partial credit and one got 0.
- Who got what and why? Describe your solution strategy and give detailed explanations.
- b. Ignoring above results, under what conditions will the three methods be identical?
- c. Give your forecast using adjusted exponential smoothing, with α = 4, β = 0.2, and T2 = 0, and plot it with original data.
Summary
The question belongs to Statistics and discusses about calculation of forecasting results with weighted moving averages, exponential smoothing, etc.
Total Word Count 339
Price: 6.00
OP: 12.00