Time Series Forecasting

Set of evenly spaced numerical data

Obtained by observing response variable at regular time

periods

Forecast based only on past values, no other

variables important

Assumes that factors influencing past and present will

continue influence in future

Example
Year: 19931994199519961997
Sales: 78.7 63.5 89.7 93.2 92.1

Time Series Components

Trend Component
Observation increase or decrease
regularly through time
Persistent, overall upward or downward
pattern
Changes due to population,
technology, age, culture, etc.
Typically several years duration

Seasonal Component
Regular pattern that repeat itself after a
period of days, weeks or months.
Due to weather, customs, etc.
Number of
Occurs within a single year
Period
Length
Seasons
Week
Month
Month
Year
Year
Year

Day
Week
Day
Quarter
Month
Week

7
4-4.5
28-31
4
12
52

Cyclical Component
Repeating up and down movements
Affected by business cycle, political, and
economic factors
Multiple years duration
Often causal or associative
relationships
0

10

15

20

Random Component
Unpredictable component that gives series
of graph irregular.
Erratic, unsystematic, residual fluctuations
Due to random variation or unforeseen
events
Short duration
and nonrepeating
M

Components of Demand

Naive Approach
Assumes demand in next period is the same as
demand in most recent period
e.g., If January sales were 68, then February sales will
be 68

Sometimes cost effective and efficient

Can be good starting point

Techniques For Averaging

Moving Average
Weighted Moving average
Exponential Smoothing

Moving Average Method

MA is a series of arithmetic means
Used if little or no trend, seasonal, and cyclical patterns
Used often for smoothing
Provides overall impression of data over time

Demand in Previous n Periods

MA
n

Where n is the number of the periods in moving average

Moving Average Example

Month

Actual
Shed Sales

3-Month
Moving Average

January
February
March
April
May
June
July

10
12
13
16
19
23
26

(10 + 12 + 13)/3 = 11 2/3

(12 + 13 + 16)/3 = 13 2/3
(13 + 16 + 19)/3 = 16
(16 + 19 + 23)/3 = 19 1/3

Potential Problems With Moving

Average
Increasing n makes forecast less
sensitive to changes
Do not forecast trend well due to the
delay between actual outcome and
forecast
Difficult to trace seasonal and
cyclical patterns
Require much historical data
Weighted MA may perform better

Weighted Moving Average

Used when some trend might be
present
Older data usually less important

Weights based on experience and

intuition

Exponential Smoothing
Form of weighted moving average
Weights decline exponentially
Most recent data weighted most

Requires smoothing constant ()

Ranges from 0 to 1
Subjectively chosen

Involves little record keeping of past data

Exponential Smoothing

w forecast = Last periods forecast

+ (Last periods actual demand
Last periods forecast)
Ft = Ft 1 + (At 1 - Ft 1)
where

Ft = new forecast
Ft 1 = previous forecast
= smoothing (or weighting)
constant (0 1)

Correlation
How strong is the linear relationship
between the variables?
Correlation does not necessarily imply
causality!
Coefficient of correlation, r, measures
degree of association
Values range from -1 to +1

Correlation Coefficient
nxy - xy
r=
[nx2 - (x)2][ny2 - (y)2]

Correlation Coefficient
y

nxy - xy
r=
[nx2 - (x)2][ny2 - (y)2]

(a) Perfect positive x

correlation:
r = +1

(b) Positive
correlation:
0<r<1

(c) No correlation:
r=0

(d) Perfect negative x

correlation:
r = -1

Multiple Regression
Analysis
If more than one independent variable
is to be used in the model, linear
regression can be extended to
multiple regression to accommodate
several independent variables
y =^a + b x + b x
1

Computationally, this is quite

complex and generally done on
the computer