Scribd
Upload
36 views14 pages

Module1 Introduction

This document provides an introduction to time series analysis and forecasting. It defines what a time series is and provides examples of continuous and discrete time series data. It describes the key components of time series decomposition including trends, seasonality, cycles, and irregular components. Finally, it discusses the concept of stationarity and different types of stationarity.

Uploaded by

KISHORE K 20MIY0004
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
36 views14 pages

Module1 Introduction

This document provides an introduction to time series analysis and forecasting. It defines what a time series is and provides examples of continuous and discrete time series data. It describes the key components of time series decomposition including trends, seasonality, cycles, and irregular components. Finally, it discusses the concept of stationarity and different types of stationarity.

Uploaded by

KISHORE K 20MIY0004
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 14

Time Series Analysis and

Forecasting
Module - 1
Introduction
 A time series is defined as a set of observations on a
variable generated sequentially in time.
 The measurement of the variable may be made
continuously or may be made at discrete.
 Examples of continuous variables on which time series
data may be obtained are the temperature on a
chemical reactor, the level of tide at a particular site or
the amplitude of an electric signal.
 Example of discrete time series data may also be
generated from the accumulation of data over a period.
Eg., monthly sales, daily rainfall, production of a crop
over different years in a country.
Discrete Variable Graph
Continuous Variable Graph
Components of Time Series
Decomposition of Time Series
 There are four categories of decomposition of time
series. i.e.,
 The Trend Component (T)
 The Seasonal Component (S)
 The Cyclical Component (C)
 The Irregular or Random Error Component (I)
The Trend Component (T)
 This is the long term upward and downward movement
of the series due to factors that influence the mean of
the series.
 Some series may remain more or less at a constant
level.
 Sudden or frequent changes are incompatible with the
idea of the trend.
 Thus trend may be defined as a slowly changing non-
random component of a time series.
The Seasonal Component (S)
 This is a measure of the characteristic behaviour of the
series during each season (Specific interval of time) in
the period, which may be one year.
 A period may be a span of 12months and a season a
quarter (3months), or a month.
 A period may be a span of 24hours of a day and a season
a three-hourly period, etc.
 The basic cause of seasonal variation is usually either
the seasonal weather
 For eg., magnitude of air line bookings is affected by
seasons
 Or some fixed period of time during which most
activities take place,
 For eg., passenger traffic during the peak hours of a day,
sales of departmental stores during the busy hours.
 Again some fixed seasonal events, eg., Christmas
celebrations, Durga Puja, Book exhibition, Sale of
fireworks.
 Demand for electricity and other forms of energy is also
seasonal.
 Demand for any product or service which depends to
some extent on weather, temperature or fixed holidays
will tend to be seasonal in nature.
The Cyclical Component (C)
 The cyclical fluctuation means the oscillatory
movement of time series, the period of oscillation being
more than a year. One complete period is called cycle.
 The length of many cycles average about 3 to 4 years,
though some are longer than 15years.
 For eg., Five years plan
 Biological or medical experiment
 Economic / GDP
 Taxation
 National / State Budget
 Etc.,
The Irregular or random error component (I)

 There is another factor which causes the variation in


the variable under study. They are not regular
variations and are purely random or irregular. These
fluctuations are unforeseen, uncontrollable,
unpredictable, and are erratic. These forces are
earthquakes, wars, flood, famines, and any other
disasters.
Stationarity
 A time series has stationarity if a shift in time doesn’t
cause a change in the shape of the distribution. Basic
properties of the distribution like the mean , variance
and covariance are constant over time.
 Why Stationarity?
 Most forecasting methods assume that a distribution has
stationarity. For example, auto-covariance and
autocorrelations rely on the assumption of stationarity. An
absence of stationarity can cause unexpected or bizarre
behaviors, like t-ratios not following a t-distribution or
high r-squared values assigned to variables that aren’t 
correlated at all.
Types of Stationarity
 Strict stationarity means that the joint distribution of
any moments of any degree (e.g. expected values, 
variances, third order and higher moments) within the
process is never dependent on time. This definition is in
practice too strict to be used for any real-life model.

 First-order stationarity series have means that never


changes with time. Any other statistics (like
variance) can change.

 Second-order stationarity (also called weak


stationarity) time series have a constant mean, variance
and an autocovariance that doesn’t change with time.
Other statistics in the system are free to change over
time. This constrained version of strict stationarity is
very common.
 Trend-stationary models fluctuate around a 
deterministic trend (the series mean). These
deterministic trends can be linear or quadratic, but the 
amplitude (height of one oscillation) of the fluctuations
neither increases nor decreases across the series.

 Difference-stationary models are models that need one


or more differencing to become stationary

You might also like