Unit-2

Dr. K. Subrahmanyam
Professor in CSE
 Statistical Modelling and Analysis
• Probability distributions
• Variations
 Probability distributions
• Probability is defined as the chance of occurrence
of an event
• Probability distribution can be classified into
discrete and continuous.
• Discrete probability distribution
– Poisson distribution, binomial distribution, tossing 3
coins, etc..
• Continuous Probability distribution
– Exponential distribution, Uniform distribution, Normal
distribution and t-distribution
 Variation
• The variation (spread) is a measure of the
extent to which the individual observations
vary with reference to the average of the
given set of observations.
• Different measures of variation are
– Range
– Quartile deviation
– Mean deviation (average deviation)
– Standard deviation
– Coefficient of variation
 Variation
Range - The difference between the highest
value (H) and the lowest value (L) in
the given set of observations

R = H-L
Coefficient of range = (H-L)/ (H+L)
 Variation
• Quartile deviation is the half of the difference
between 3rd quartile and the 1st quartile.
• Mean deviation is the mean of the absolute
deviations of the individual observations from
either arithmetic mean or median of the given
set of observations.
 Variation
• Standard deviation ( root mean square
deviation) is widely used measure of variation.

• xi is the ith observation, x is mean of

observations, N the number of observations.
 Variation
• Coefficient of Variance = (SD/Mean) *100.
 Multivariate Analysis
• Multivariate Analysis is the analysis of data
consisting of observations pertaining to a
reality involving 2 or more variables.
– Basic topics of MVA
• Correlation Analysis
• Forecasting
• Linear and Non-Linear Regression
• Time series
Difference between Correlation and
Regression
• Correlation analysis is concern with
knowing whether there is a relationship
between variables.
• Regression analysis is concern with finding
a formula that represents the relationship
between variables so as to find an
approximate value of one variable from the
value of the other(s)
 Hypothesis testing procedure
• Suppose there is a claim or a statement.
• The null hypothesis is an answer in which the claim
or the statement is true, and denoted by H0.
• The alternative hypothesis is an answer in which the
claim or the statement is NOT true, and denoted by
Ha (HA or H1).
– If we conclude H0, we say we accept H0 or fail to reject H0
– if we conclude Ha, we say we reject H0 and conclude Ha ".
 A testing procedure is specified by the following steps
• Choose a test statistic. A test statistic is a function of
data which makes the decision about the rejection or
acceptance of H0.
• A rejection range is a set in which we reject H0 if the
test statistic is in the set and accept H0 if the test
statistic is outside of the set.
• Let T be the test statistic and R be the re-jection
range. Then, we conclude
• H0 if T is not member of R
• Ha if T is member of R
 Time series

• The term time series means a set of

observations concurring any activity against
different periods of time. In order to describe
this flow of economic activity, the statistician
uses a time series.
• Examples-
– Number of faculty recruited by KLU for each of the past 5 years
– Number of fatal road accidents in Andhra Pradesh for each day for the
past 3 months.

http://220.227.161.86/21538sm_finalnew_vol2_cp19.pdf
 Time series

• Time series analysis comprises methods for

analyzing time series data in order to extract
meaningful statistics and other characteristics
of the data

http://220.227.161.86/21538sm_finalnew_vol2_cp19.pdf
 Components of a time series
• Any time series can contain some or all of the
following components:
1. Trend (T)
2. Cyclical (C)
3. Seasonal (S)
4. Irregular (I)
• These components may be combined in different ways. It is
usually assumed that they are multiplied or added, i.e.,
– yt = T x C x S x I
– yt = T + C + S + I
 Components of a time series
Trend component
The trend is the long term pattern of a time series. A trend can be positive or negative
depending on whether the time series exhibits an increasing long term pattern or a decreasing
long term pattern. If a time series does not show an increasing or decreasing pattern then the
series is stationary in the mean.

Cyclical component
Any pattern showing an up and down movement around a given trend is idenfied as a
cyclical pattern. The duration of a cycle depends on the type of business or industry
being analyzed.

Seasonal component
Seasonality occurs when the time series exhibits regular fluctuations during the same month
(or months) every year, or during the same quarter every year. For instance, retail sales peak
during the month of December.

Irregular component
This component is unpredictable. Every time series has some unpredictable component that makes it a
random variable. In prediction, the objective is to \model" all the components to the point that the only
component that remains unexplained is the random component.
Methods of fitting a straight line to a time series

1.Free hand method

2.Method of semi averages
3.Method of moving averages
4.Method of least squares
Example Method of Least Squares

http://220.227.161.86/21538sm_finalnew_
vol2_cp19.pdf
 Time series Analysis
• 8 Steps in time series Analysis:
1. Input the original time-series data (y)
2. Find moving Averages (MA)
3. Obtain seasonal indices(S)
4. Obtain deseasonalised value (Y) of y by using the formula
Y = y/S
5. Fit a regression model for the depersonalised forecast Y =
a+bX
6. Obtain trend estimates (Y) using Y = a+bX
7. Find the cyclical component using the equation C= Y/Y
8. Stop.
Note : The cyclical component cannot be used for estimation of
forecast, but is used only for Analysis. (Y is Y cap)
 Error Analysis
• Types of errors
 Error Analysis
• Standard measures of errors
– Absolute Deviation
– Relative % Deviation
– Standard Deviation
 Error Analysis
• Propagation of Errors
 Spectral Analysis

• Digital Signal Processing (DSP) has revolutionized spectral analysis. Where the
swept spectrum analyzer dominated the market in the past, the Fast Fourier
Transform (FFT) based spectrum analyzer is now gaining acceptance as the
method of choice. This is due in part to the prevalence of high speed, high
dynamic range Analog-to-Digital Converters (ADC) and high speed signal
processing devices such as Field Programmable Gate Arrays (FPGA). Because the
FFT-based spectrum analyzer is readily implemented with a limited set of generic
hardware, it is an attractive technique for Synthetic Instruments (SI), where the
goal is to form multiple measurement functions from a limited set of generic
hardware modules. In this paper some of the fundamental design parameters and
specifications of modern spectrum analyzers such as dynamic range,
instantaneous bandwidth, and image rejection are presented. These parameters
are explored with a focus on maintaining system performance without sacrificing
flexibility.