Page 29 - IMDR JOURNAL 2020-21

IMDR’s Journal of Management Development and Research 2020-21

PYTHON BASED MARKOV CHAIN
APPROACH FOR CONSUMER

AND MARKET ANALYSIS

Dr. Mrs. P.N. Gokhale, Dr. S.M. Bakre, Dr. S.M. Dashputre
1Professor and HOD, Department of Electrical Engineering, JSPM JSCOE, Hadapasar, Pune
drpngokhalejscoe@gmail.com
2Associate Professor, Department of Electrical Engineering, AISSMS IOIT, Pune
shashikant.bakre@aissmsioit.org
3Ex. Director, IST Institute of Management, Pune
istpune@yahoo.co.in

Abstract- limiting distributions. Multiple states may arise
Over the period of time, the consumers and during the modification process of Markov Chain.
markets are becoming more and more complex The Markov Chain essentially comprises
day by day and it has been a challenging task to of Transition matrices based on probabilistic
analyze them. The Markov Chain is one of the distribution. This can be explained with the climatic
traditional techniques that can be effectively applied conditions forecasted by two agencies let A and B. As
for analyzing consumers and markets. However the per the records of agency A there is 80% probability
main constraint in implementing Markov Chain is to be a sunny day whereas the observations of agency
preparing Markov Model and performing operations B predict 70% chances of Sunny day. Therefore the
on consumer and transition matrices based on huge prediction of 80% possibility of sunny day indicates
and complex data. In the advent of upcoming that balance 20% probability would be having a
technologies related to Data Science, it has become cloudy day. Similarly in view of agency B, there are
easier to perform these tasks. The paper presents 30% chances of being a cloudy day. This indicates
a novice method of conducting consumer and that balance 70% chances are in favor of sunny day.
market analysis using Python based Markov Chain This relationship between sunny and cloudy days
approach. The sample source code and algorithm are can be expressed in Transition Model illustrated in
furnished with a particular example. The proposed Fig 1.
paper supports Atma-Nirbhar concept announced
by the Government of India under pillar ‘demand’.

I INTRODUCTION
The Markov Chain is a stochastic model that
describes the sequence of probable events in which
the probability of each event depends only on the
state achieved by the current event. It is commonly Fig. 1. Transition Model
used, relatively simple, intuitive and accessible Based on transition model developed as shown
method for ethnological development. in Fig 1, the transition matrix can be prepared as
This technique was developed by the Russian follows. The matrix can also be expressed in terms
mathematician Andrei A. Markov (1856-1922) in of numbers by dividing each matrix element by 100.
the year 1906.He developed a mathematical model It should be noted here that 80% and 30% are the
that illustrates a sequence of possible events such that base records maintained by the agencies A and B
the probability of each event is dependent on current respectively from which balance percentages of 20%
event. The Markov chains are used in various areas and 30% have been worked out respectively.
such as finance, stock markets, census measurements,
marketing and supply chain management. As a
stochastic process, Markov Chain has properties of
reducibility, periodicity, steady state analysis and

Page No | 27