Example of simple exponential smoothing, Mechanical Engineering

Example of simple exponential smoothing:

A TV manufacturer utilizes simple exponential smoothing with α = 0.1 to forecast demand. For the first week of March, the forecast was 400 units while actual demand turned out to be 370 units.

1. Forecast the demand for the second week of March.

2. If the actual demand for the second week turned out to be 417 units, forecast the demand for the 3rd week. Continue forecasting for the subsequent four weeks. You can take actual demands for the subsequent weeks as 409, 388, 377, 445, and 410 units.

3. Develop an adjusted exponential forecast for this company. You may assume the initial trend adjustment factor Tt as zero and β = 0.1.

Solution

(a) Ft + 1 = Ft + α (Dt - Ft)

                     = 400 + 0.1 (370 - 400) = 397 units.

 (b)

707_Adjusted Exponential Smoothing.png

(c)        The trend adjustment is accomplished by total of a correction factor

[(( 1 -β )/ β ) Tt +1 )]   to the simple exponential forecast.

As the initial trend adjustment factor (given) Tt = 0, we have

March 1st Week

Tt + 1 = β (Ft + 1 - Ft) + (1 - β) Tt

= 0.1 (397 - 400) + (0.9) (0) = - 0.30

T(t +1) adj = Ft +1 +( 1 -β / β )T t +1

= 397 +( 0.9/0.1)  (- 0.30) = 394.30

The adjusted forecast for the second week of March = 394.30.

March 2nd Week

Tt + 1 = 0.1 (399 - 397) + (0.9) (- 0.30)

= 0.2 - 0.27 = - 0.07

F(t + 1) adj = 399 + 9 (- 0.07) = 398.37

The remaining values may be put in table form, as below.

606_Adjusted Exponential Smoothing1.png

The trend-adjusted forecast for the third week of April is Ft, adj   = 393.54 units whereas simple exponential forecast for the same week is 400 units.

The main reason for broad use of simple and adjusted exponential smoothing is that the forecast may be computerized quickly, requires very little memory space, and data can be updated easily. These methods are specifically useful to firms which have thousands of items in inventory. Since each forecast value contains all past forecast and demand data in the properly exponentially weighted manner, there is no requirement to maintain records of historical data (as in case of arithmetic moving averages). It is an effective and efficient method of forecasting with a built-in means of tracking the average while discounting the erratic random fluctuations.

Posted Date: 2/1/2013 12:05:04 AM | Location : United States







Related Discussions:- Example of simple exponential smoothing, Assignment Help, Ask Question on Example of simple exponential smoothing, Get Answer, Expert's Help, Example of simple exponential smoothing Discussions

Write discussion on Example of simple exponential smoothing
Your posts are moderated
Related Questions
#tool life formular for diferent n values

With the help of diagram , depict the evaluation of Machine Design. Discuss in detail Technical feasibility, finanacial and economic feasibility, social and environmental feasib

Eccentricity : This is distance between the axis of rotation and CG of mass. Unbalance : This is the product of mass and eccentricity. Balancing : This is

Q. Explain Double Bolting? Where low strength bolts must be used, initial bolting with higher strength low temperature bolts may be necessary to minimize accumulative strain an

Explain the Cost analysis of radial drilling machine Cost regarding to Manual arrangement of auger extension in paver machine: 1.cost of segmented angle radial drill machin

necessity of lubrication

Calculate stress intensity factor: For the plate in Figure (a) W = 25 mm, 2a = 10 mm, plate thickness, t = 2 mm, Load P = 1000 N. Calculate stress intensity factor if length o

Evolution of Safety - Industrial Safety The large scale expansion started taking shape along with the beginning of nineteenth century but safety could not turn into the concer

Water flows into the sink shown in figure 5 at a flow rate of 8 L/min. If the drain is closed, the water will eventually flow through the overflow drain holes rather than over the

(a) Illustrate reversible, irreversible and Quasi-static process with examples. (b) What do you mean by terms change of state, path, process, property, cycle. (c) Distinguish