Scribd
Upload
4 views

Lesson3

Uploaded by

rajithsjc
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
4 views

Lesson3

Uploaded by

rajithsjc
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 13

Time Series Models

LESSON 3
How to choose an appropriate model for
given time series?
• Can we assume time series data are iid?
• Usually, statistical theories depend on the assumption iid not true for time order observation.
• When successive observations are dependent, future values may be predicted from past
observations.
• For analysis and modeling purposes we are interested in whether or not consecutive
observations are related in some way. If so, we might be able to use previous values to help
forecast the next value.
• One graphical way to investigate that question is to pair up consecutive values and plot the
resulting scatterplot of pairs.
Example-Pharmaceutical product sales and
Lag one Correlations
Example 2-Chemical process viscosity readings
Stationary Time Series Models
Stationary Time Series Models ctd.
• This implies, If 𝑋 is strictly stationary, E[𝑋 ] is a constant in t. Also, cov(𝑋 , 𝑋 ) cannot
depend on time.
• If we have n random variables 𝑌 , 𝑌 , … , 𝑌 , the joint distribution would tell us everything.
This is seldom possible.
• Thus the version of stationarity in above definition is too strong for most applications.
Moreover, it is difficult to assess strict stationarity from a single data set. Rather than
imposing conditions on all possible distributions of a time series, we will use a milder version
that imposes conditions only on the first two moments of the series. We now have the
following definition.
Weakly Stationary Series

• Speaking stationary means there is no systematic change in mean and no systematic


change in variance.

Strictly stationary =) Weakly stationary.


What about the converse?
Autocovariance Function
Autocorrelation Function

• The autocorrelation measures the correlation between an observation at time t, Xt and


observation at time t + h, Xt+h.
• Since this is the correlation between observations in the same series at different time
periods, it is called autocorrelation.
Basic Properties of the Autocovariance
Function
Examples of Time Series Models
Sample Statistics
• Suppose that we have observed a time series 𝑥 , 𝑥 , … . , 𝑥 , .
• To assess the degree of dependence in the data and to select a model for the data that reflects
this, one of the important tools we use is the sample autocorrelation function (sample ACF) of
the data.
• If we believe that the data are realized values of a stationary time series 𝑋 , then the
sample ACF will provide us with an estimate of the ACF of 𝑋 .
• This estimate may suggest which of the many possible stationary time series models is a
suitable candidate for representing the dependence in the data.
Sample Statistics
𝑋 is stationary with mean µ and ACVF γ(·). For observed values of a series, 𝑥 , 𝑥 , … . , 𝑥 ,
estimates for mean, sample ACVF and sample ACF are

You might also like