MSTE-002

Industrial Statistics-II
Indira Gandhi National Open University
School of Sciences

Block

4
TIME SERIES MODELLING
UNIT 13
Trend Component Analysis 5
UNIT 14
Seasonal Component Analysis 33
UNIT 15
Stationary Processes 55
UNIT 16
Time Series Models 73
Curriculum and Course Design Committee
Prof. K. R. Srivathsan Prof. Rahul Roy
Former Pro-Vice Chancellor Theoretical Statistics and Mathematics Unit
IGNOU, New Delhi Indian Statistical Institute, New Delhi

Prof. Parvin Sinclair Prof. Diwakar Shukla

Former Pro-Vice Chancellor Department of Mathematics and Statistics
IGNOU, New Delhi Dr. Hari Singh Gaur University, Sagar

Prof. Geeta Kaicker Prof. Rakesh Srivastava

Former Director, School of Sciences Department of Statistics
IGNOU, New Delhi M.S. University of Baroda, Vadodara

Prof. Jagdish Prasad Prof. G. N. Singh

Department of Statistics Department of Applied Mathematics
University of Rajasthan, Jaipur I.S.M., Dhanbad

Prof. R. M. Pandey Prof. Gulshan Lal Taneja

Department of Bio-Statistics Department of Mathematics
AIIMS, New Delhi M.D. University, Rohtak

Dr. Manish Trivedi Prof. Poornima Mital

School of Sciences, IGNOU School of Sciences, IGNOU

Dr. Nitin Gupta Prof. Sujatha Varma

School of Sciences, IGNOU School of Sciences, IGNOU

Dr. Neha Garg Dr. S. Venkataraman

School of Sciences, IGNOU School of Sciences, IGNOU

Mr. Rajesh Kaliraman Dr. Deepika

School of Sciences, IGNOU School of Sciences, IGNOU

Block Preparation Team

Prof. Rakesh Srivastava (Editor) Prof. G. K. Shukla (Units 13 to 16)
Department of Statistics (Retd.), Department of Applied Mathematics
M.S. University of Baroda Indian Institute of Technology
Vadodara (GUJ) Kanpur (UP)

Prof. Vijayshri (Language Editor) Dr. Manish Trivedi (Units 13 to 16)

School of Sciences, IGNOU School of Sciences, IGNOU

Programme and Course Coordinator: Dr. Manish Trivedi

Production
Mr. Sunil Kumar, AR (P)
School of Sciences, IGNOU

September, 2014
© Indira Gandhi National Open University, 2014
ISBN-978-81-266-
All rights reserved. No part of this work may be reproduced in any form, by mimeograph or any other
means, without permission in writing from the Indira Gandhi National Open University
Further information on the Indira Gandhi National Open University may be obtained from the
University’s Office at Maidan Garhi, New Delhi-110068 or visit University’s website
http://www.ignou.ac.in
Printed and published on behalf of the Indira Gandhi National Open University, New Delhi by the
Director, School of Sciences.
Laser Typeset by: Tessa Media & Computers, C-206, A.F.E.-II, Okhla, New Delhi
Printed at:
BLOCK 4 TIME SERIES MODELLING
In the previous block, we have discussed the fundamental concepts of
regression modelling. You have learnt that regression analysis is a statistical
tool for investigating and analysing the average relationship between two or
more variables. We have also discussed the applications of regression analysis
in decision making. In this block, we discuss time series modelling, which
helps us in analysing past data to forecast and plan for future. For example,
suppose the Government wishes to bring in a policy to ensure low cost housing
for economically weaker sections (EWS). Then it would like to know: What
would the number of people in this category be in the next five years? In
general, institutions, organisations, governments and even individuals take
most decisions based on the situations expected in future. Planning for future is
an essential aspect of decision making and managing any organisation. This
requires that one should be able to forecast the future demands or trends.
For this, it is natural to use historical data of the last few years. Based on the
features observed from the past data, we try to understand their role in causing
variability and use them for forecasting. This is done with the help of time
series modelling, which is the analysis of a collection of observations made
sequentially through time to forecast future trends. We discuss basic concepts
of time series modelling in the last unit of this block. It comprises four units.
In Unit 13 entitled Trend Component Analysis, we introduce the concept of
time series. We discuss different types of time series and different components
of time series along with examples. We also explain the methods of smoothing
and filtering the time series data and estimation of trend by curve fitting
method and curvilinear method.
When time series data do not exhibit any trend and cyclic components but
reflect seasonal variations, we have to estimate the seasonal component by
removing irregular components. In Unit 14 entitled Seasonal Component
Analysis, we discuss some methods for estimating the seasonal components,
namely, the simple average method, ratio to moving average and ratio to trend
methods.
In Unit 15 entitled Stationary Processes, we describe a very important class of
time series called the stationary time series. If a time series shows a particular
type of non-stationarity then some simple transformation makes it stationary
and we can model it. We also explain the concepts of autocovariance,
autocorrelation and correlogram for a stationary time series.
In Unit 16 entitled Time Series Models, we discuss the Autoregressive (AR),
Moving Average (MA), mixed Autoregressive Moving Average (ARMA) and
Autoregressive Integrated Moving Average (ARIMA) processes. We also
analyse their properties in the form of autocorrelations and discuss the fitting of
suitable models to the given data.
 Notations and Symbols
Yt : Time series value at time t of variable of interest Y
Xt : tth observation of the independent variable X
Ŷt : tth observation of the fitted value
et : tth forecast error
Tt : Trend component or trend effect
St : Seasonal component or seasonal effect
Ct : Cyclic component or cyclic effect
It : Irregular component or irregular effect
Si : ith seasonal index
yi : Average for ith year, quarter or month
µ : Mean
at : tth white noise with mean zero, variance  2a
ψi : ith weight
βi : ith constant
k : Autocovariance at lag k
ck : Autocovariance coefficient at lag k
ρk : Autocorrelation at lag k
rk : Autocorrelation coefficient at lag k
B : Backward shift operator
ρ = a.c.f. : Autocorrelation function
α = p.a.c.f. : Partial autocorrelation function
MA(q) : Moving Average process of order q
AR (p) : Autoregressive process of order p
ARMA(p, q) : Autoregressive Moving Average process of order (p, q)
ARIMA(p, d, q) : Autoregressive Integrated Moving Average process of
order (p, q) with d differences
{Y(t)}=Yt : Descrete time random process
V{Y(t)} : Variance of time dependent random variable Y
Sec. : Section/subsection
Secs. : Sections/subsections
Fig. : Figure
Figs. : Figures