Free Management Dissertations - 1. Tabulation Of Demand And Average Demand The Actual Demand, 3-point Moving
1. Tabulation of demand and average demand
The actual demand, 3-point moving average (without trend adjustment) and exponentially smoothed average (with smoothing constant 0.3) are tabulated for the year. .
January projections are calculated by adding the change in the moving average from November to December to the December moving average. (Table 1)
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Jan
Demand
23
19
26
31
29
30
38
40
29
53
57
70
3-point MA
23
25
29
30
32
36
36
41
46
60
74
Exp. Smoothed MA
22
23
25
26
27
30
33
32
38
44
52
60
Table 1
A.2 Plot of moving averages and actual demand (Figure 1)
Figure 1
A.3 Inferences from graphs of actual and smoothed demand
All three plots show an accelerating trend of demand. The two moving averages both underestimate the actual demand, exceeding actual demand in February and September only. This is principally because moving averages are based on history and therefore in the absence of trend correction will always lag behind the curve of best fit through the actual demand.
A.4 Mean error for each of the two forecasting methods
The sum of the errors (Demand Forecast) over 10 periods for 3-point moving average is 45
Mean error (3 point MA) = 45/10
= 4.5 devices per period
The sum of the errors (Demand Forecast) over 11 periods for exponential smoothed average is 70 devices
Mean error (exponential smoothed average) = 70/11
= 6.4 devices per period
Using the absolute error for each gives MAD’s of 5.9 and 7.2 respectively for the two forecast methods.
A.5 Evaluation of forecasting methods
The 3-point moving average appears the better forecasting method, simply because it is weighted more closely to the most recent demand than is the exponentially smoothed average with the factor given. It therefore lags less than the latter, and gives a closer approximation under sustained trend. This would not necessarily be the case for fluctuating demand with a less significant trend. If these forecasts are to be used for capacity planning, trend correction should be applied along with known external factors (if any) limiting demand.
B.1 Tabulation of cumulative demand and capacity for the fitting operation
Cumulative demand (in terms of arrivals) and cumulative capacity (in terms of departures) are tabulated for the year, for the fitting operation (Table 2). Note that the supply (departures) in any period in the fitting operation cannot exceed the number of patients, and therefore may be less than the clinic capacity.
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
Dec
Cum. Demand
23
42
68
99
128
158
196
236
265
318
375
445
Cum Supply
20
42
68
92
120
148
176
232
265
313
369
425
Backorders
3
0
0
7
8
10
20
4
0
5
6
20
Table 2
Figure 2
B.
Dissertations - Free Management Dissertations
