Split/Skill Forecast Methods - Lucent Technologies CentreVu Release 3 Version 5 Forecast User Manual

Introduction
Getting Started with Forecasts
Split/Skill Forecast
Methods
One-week-apart data points are days in each of the preceding 4
weeks and are the same day of the week as the forecasted day.
Look at the example in
June 5, and today is Wednesday, May 28, the data points would be
the Thursdays on May 1, 18, 15, and 22.
S M
4 5
11 12
18 19
25 26
= today
Figure 1-2: Sample Data Points—One Week Apart
For some types of split/skill forecasts, you specify at run time which type
of data points CentreVu CMS should use. For the Current Day and
Intraday forecasts, CentreVu CMS automatically uses 1-week-apart data
points.
The Special Day Forecast uses data from only the last
Note
occurrence of a special day, not from data points.
CentreVu CMS uses the data from the data points in different ways,
depending on which of the four forecast methods you select. The four
1
available forecast methods are as follows:
No Trending
Current Trending
CentreVu CMS R3V5 Forecast 585-215-825
Figure 1-2. If the forecasted day is Thursday,
MAY 1997
T W T F S
1 2 3
6 7 8 9 10
15
13 14 15 16 17
22
20 21 22 23 24
29
27 28 29 30 31
= data points
CentreVu CMS finds the average number of
calls, per interval, in the data points.
CentreVu CMS finds the trend of the number of
calls in the data points. Thus, if the number of
calls in the most recent data points indicates
that the number of calls is going down, the
calls forecasted will be lower than they would
have been forecasted with the "no trending"
method.
JUNE 1997
S M T W T F
3 4 5 6
1 2
8 9 10 11 12 13
16 17 18 19 20
23 24 25 26 27
30
= forecast day
1-6
S
7
14
21
28