# Small questions on Forecasting

Part B - You decide to use more sophisticated methods to forecast future sales. Prepare a linear regression and exponential smoothing forecast (with alpha = 0.2) for the data (assume that the initial forecast is equal to the average of all demand data). Calculate the mean absolute deviation (MAD) and tracking signal (TS) for each forecast.

Demand

April, 2011 22000

May, 2011 25,080

June, 2011 24,578

July, 2011 24,333

August, 2011 26,036

September, 2011 25,515

October, 2011 28,322

November, 2011 29,172

December, 2011 32,089

January, 2012 36,902

February, 2012 38,747

March, 2012 37,197

"Linear

Regression" "Exponential

Smoothing" Demand

1. What is the mean absolute deviation (MAD) through March, 2012 for each method?

2. What is the tracking signal (TS) through March, 2012 for each method? 3. 3. What is the linear regression forecast for April, 2012? . 4. What is the exponential smoothing forecast for April, 2012? 5. 5. Which method provides the best forecast (refer to part A for the moving average and weighted moving average forecast information)?

6. What is the mean absolute deviation (MAD) through March, 2012 for each method? linear Regression "Exponential

Smoothing"

a)1300 A)4000

b)1400 b)5000

c)1500 c)6000

d)1600 d)7000

2. What is the tracking signal (TS) through March, 2012 for each method?

linear Regression "Exponential

Smoothing"

a)-1 a)2.5

b)0 b)3.5

c)1 c)4.5

d)2 d)5.5

3. What is the linear regression forecast for April, 2012?

a)38000

b)38300

c)38600

d)38900

4. What is the exponential smoothing forecast for April, 2012?

a)32750

b)33,000

c)33,250

d)33,500

5. Which method provides the best forecast (refer to part A for the moving average and weighted moving average forecast information)?

Part D - You have narrowed your forecasting process down to an exponential smoothing method. You need to evaluate the simple exponential smoothing method (without trend) and compare it to the trend-adjusted exponential smoothing method to finalize your decision. Using the data shown below, prepare a forecast using each method and calculate the Cumulative Forecast Error (CFE) and Mean Absolute Deviation (MAD) to determine which method is more appropriate. Use the indicated values for alpha and beta for your forecasting process. The initial forecast (and trend where needed) are given with the data.

1. What is the Month 13 forecast for each method?

2. What is the total Cumulative Forecast Error (CFE) for each method?

3. What is the Mean Absolute Deviation (MAD) for each method?

4. Which method is better, and why?

5. If the value of alpha (?) were changed to 0.1 for both methods, which method would be better, and why?

Month Demand Forecast

1 2000 2,000

2 1940

3 1785

4 1518

5 1412

6 1271

7 1157

8 1111

9 912

10 885

11 744

12 707

Trend Adjusted Exponential Smoothing

Alpha (?) = 0.3

Beta (?) - 0.1

Month Demand Forecast Trend

1 2000 1800 200

2 1940

3 1785

4 1518

5 1412

6 1271

7 1157

8 1111

9 912

10 885

11 744

12 707

1. What is the Month 13 forecast for each method?

Simple Exponential Trend Adjusted

Smoothing Exponential Smoothing

a)900 a)900

b)1030 b)1030

c)1160 c)1160

d)1290 d)1290

2. What is the total Cumulative Forecast Error (CFE) for each method?

Simple Exponential Trend Adjusted

Smoothing Exponential Smoothing

a)5000 a)6200

b)4300 b)5800

c)3600 c)5400

d)2900 d)5000

3. What is the Mean Absolute Deviation (MAD) for each method?

Simple Exponential Trend Adjusted

Smoothing Exponential Smoothing

a)300 a)300

B)390 B)390

c)480 c)480

d)570 d)570

4. Which method is better, and why?

5. If the value of alpha (?) were changed to 0.1 for both methods, which method would be better, and why?

which method would be better, and why

© BrainMass Inc. brainmass.com June 4, 2020, 2:37 am ad1c9bdddfhttps://brainmass.com/business/business-management/small-questions-on-forecasting-470201

#### Solution Summary

This posting containis 2 small forecasting questions which involve forecasting using Regression, Exponential Smoothing and Exponential smoothing with trend and calculation of MAD and tracking signal for each of the forecast.