DATA SCIENCE APPROACH TO STOCK MARKET ANALYSIS

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 10 Issue: 08 | Aug 2023 www.irjet.net p-ISSN: 2395-0072
© 2023, IRJET | Impact Factor value: 8.226 | ISO 9001:2008 Certified Journal | Page 316
DATA SCIENCE APPROACH TO STOCK MARKET ANALYSIS
Sneh Patil1, Yash Bharat Jadhav2, Neel Gude3, Anish Shivram Gawde4, Mohammed Naif5, Aditi
Bombe6
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Stock analysis is atechniquethataidsbuyersand
sellers for investors and traders. Investorsandtradersattempt
to obtain an advantage in the markets by making educated
decisions by researching and assessing both historical and
recent data. We only care about the information that is
pertinent to us in the information-overflowing world. The
material can be categorised according to a number of
disciplines, including engineering, the arts, science, history,
sports, geography, and economics. In these circumstances, itis
crucial for us to focus on the information that is pertinent to
us.[16] The information that is provided is poorly organised
and only provides information on the current situation,
particularly when it comes to the sector of finance. On the
stock market, virtually all significant economic transactions
take place at a dynamic rate known as the stock value based
on market equilibrium. A precise forecast of this stocktrend in
advance could result in enormous rewards.
Key Words: Data Science, Stock Market, Algorithms,
Trade, Data Analysis
1.INTRODUCTION
The stock market is quite complicated and prone to
volatility, and it is influenced by a number of factors. Due to
the market's unpredictability, investors can benefit or lose
money. Stock analysis is a tool used by traders and investors
to assist them decide what to purchase and sell. Investors
and traders strive to obtain an advantage in the markets by
making educated decisions by researching and analysing
both historical and recent data. A precise forecast of this
stock trend in advance could result in enormous rewards.
[14, 21] Two strategies are used in this paper to achieve the
prediction. Utilising different technical indicators is one of
them. These technical indicators are used to predict
upcoming shifts in stock trend. [21] The technical indicators
used in this stock market forecasting tool will substantially
support the predictability.
The second strategy, which is based on the Hidden Markov
Model, is more probabilistic. This model has been broad and
is more suited for dynamic systems. used when solving
pattern recognition issues. [22, 23] It isfocusedonstatistical
techniques and has a robust capacity for handling new data.
Based on historical datasets that complement current stock
price behaviour, this model interpolates the two datasets
using suitable neighbouring price elements.[16, 15] As a
result, a prediction is made regarding the variable of
interest's stock trend tomorrow. [22] Fundamental analysis,
which considers the financial performance of the specific
firm we are analysing as well as significant financial news
regarding the company, can be used to analyse the stock
market. [14] This leads us to news analysis, where we must
parse through all the pertinent company newsandpapers to
obtain the necessary data so that we can decide whether or
not to invest. The "Theory of Demand and Supply" underlies
the functioning of the stock market [12, 18]. It asserts that
supply and demand are always negativelycorrelated.[18]As
a result, as demand rises, purchasers seize control of the
stock market, which then turns bullish. On the other side, as
soon as supply enters the market, sellers seize control and
stock prices begin to decline sharply. The "Head and
Shoulder Theory" [18], which states that the highest priceof
any stock is considered the head and the next peak high is
called the shoulder, can also be used to analyse the stock
market. The majority of the indicatorswe'll encounterinthis
study have a statistical basis. As a result, each technical
indicator's design step involves some assumptions.
2. UNDERSTANDING STOCK MARKET
Although practically anything can now be purchased online,
each commodity often has its own market. For instance,
individuals travel to farms and the outskirts of cities to
acquire Christmas trees, go to the neighbourhood timber
market to buy wood and other supplies for house
renovations, and shop at Walmartfortheirweeklygroceries.
Such a focused marketplace acts as a venue for various
buyers and sellers to connect, communicate, and conduct
business. One can be sure that the pricing is fair because
there are so many market participants. For instance, if there
is only one Christmas tree vendor in the entire city, he will
be free to charge whatever he wants because there will be
nowhere else for customers to go. If there are many tree
sellers in a single marketplace, they will have to compete
with one another to drawcustomerstopurchasetheirgoods.
It will be a fair market with transparent pricing because
purchasers will have a wide range of options. Even when
buying online, customers compare prices from several
merchants on the same app or across different apps or
websites to find the greatest offers, which compels the
different online vendors to provide them the best deal. A
stock market is a comparable venue set aside for the trading
of various securities in a regulated, secure, and controllable
setting. The stock market ensures fair pricing practises and
transparency in transactions by bringing togetherhundreds
of thousands of market players who want to purchase and
sell shares of various companies. Modern computer-aided
stock markets operate electronically, which makes them
more and more advanced. Earlier stock markets used to

work on the issue and transactinpaper-basedphysical share
certificates.
In short, stock markets offer traders a safe, regulated
environment where they can confidently trade shares and
other permissible financial products with little to no
operational risk. The stock marketsfunctionasbothprimary
and secondary markets while adhering to the regulator's
established guidelines. The stock market, which serves as a
primary market, enables businesses to issue and sell their
shares to the general public for the first time through the
initial public offering (IPO) procedure. This practise aids
businesses in obtaining the funding they want from
investors. It basically implies that a business divides itself
into a certain number of shares, such as 20 million, and sells
a certain portion of those shares, such as 5 million, to the
general public at a certain price, such as $10 a share. A
corporation needs a market where these shares can be sold
to make this process easier. The stock market provides this
marketplace. If everything goes according to plan, the
corporation will sell the 5 million shares at a price of $10
each and make a profit of $50 million. In anticipation of
growing share prices and any prospective income in the
form of dividend payments, investors will receive business
shares that they can anticipate to hold for the term of their
choice. The corporation and its financial partners pay the
stock market a fee for its services as a facilitator of this
capital-raising process.
The stock exchange also acts as the trading platform that
enables routine buying and selling of the listed shares after
the first-time share issuance IPO exercise, also known asthe
listing process. The secondary market is comprised of this.
Every trade that takes place on the stock exchange's
platform during secondary market activity generates a
charge for the stock exchange.Thetask ofguaranteeingprice
transparency, liquidity, price discovery, and fair dealings in
such trading activities falls to the stock exchange. The
exchange maintains trading systems that effectivelymanage
the buy and sell orders from diverse market participants
because nearly all major stock marketsacrosstheworld now
operate electronically. In order to facilitate transaction
execution at a price that is fair to both buyers and sellers,
they carry out the price matching function. A listed firmmay
later conduct other offerings, such as follow-on offers or
rights issues, to issue new, extra shares. They might even
buy back shares or take them off the market. Such tradesare
facilitated by the stock exchange. The S&P 500 indexandthe
Nasdaq 100 index, which provide a gauge to follow the
movement of the whole market,areonlytwo examplesof the
numerous market-level and sector-specific indicators that
the stock exchange frequently develops and maintains.
The stochastic oscillatorandstochastic momentumindexare
other techniques. All corporate news, announcements, and
financial reporting are also maintained by the stock
exchanges and are typically accessible on theirsites.Various
additional corporate-level, transaction-related activitiesare
also supported by a stock market. Profitable businesses, for
instance, may reward investors by providing dividends,
which are often derived from a portion of the company's
profits. All of this information ismaintained bythe exchange,
which also has the potential to facilitate some of its
processing.
2.1 Interval
Every second, there are millions of trades made in the
market. So, in order to analyse the data, we must categorise
these trades according to the time and period in which they
occurred. These trades can be divided into short-term and
long-term time frames. One minute, two, five, ten, fifteen,
thirty, and sixty minutes are additional categories for intra-
day. Long-term intervals fall within the daily, weekly,
monthly, and so forth categories. Count the Costs Any stock
will be associated with 4 different price categories at every
interval. High, Low, Open, and Close prices are all displayed.
The greatest value at which it was exchanged during that
period of timeis what is meant by the high price. The bargain
price represents the lowest amount it went for during that
time. The price at which the stock was last traded duringthat
interval is known as the "close price," while the "open price"
refers to the first deal that occurred during that period.
2.2 Trend
Any stock's demand at any one time may be higher than
its supply or lower than its available supply, depending on
the trend. As a result, we can divide this Trend into two
categories, namely Bearish and Bullish. When there is more
demand than supply at a given period, a stock is said to be in
a bullish trend. A stock is said to be in a bearish trend if there
is a greater available supply than there is demand.
3. ANALYSIS OF STOCK MARKET
The most prevalent application of algorithmic trading is
high-frequency trading (HFT). Trading involves buying a
potential share at a discount and then selling it at the
market's top growth rate. To ascertain the stock's potential,
a thorough statistical verificationandstock researchprocess
is used. The factors that affect this trade activity include
time, price, volume, and technical indicators. The trading
choice should take human parallax errorsinto consideration
at every stage. In the case of algorithmic trading, these
actions are specifically planned to maximize profit while
lowering risk in each trading transaction. The algorithm can
process in a timely cycle, allowing it to complete more
transactions in a given time frame.
3.1 Basic Technical Trading Signals
We can assess market activity and predict future market
behaviour thanks to several crucial technological toolsofthe
trade.

 Moving Average ConvergenceDivergence(MACD): After
the trading circumstances have been created, moving
average convergence and divergence(MACD)indicators
give indications. known also as a trailing signal.
 Aroon Indicator: The Aroon Indicator tracks the new
highs and lows in the price movement of the market
trend.
 The Average Directional Index (ADI): A price moving
trend's momentum and intensity are gauged by the
Average Directional Index (ADI). The directional
strength is strong when the ADX value is more than 40.
When it drops below 20, the vigour is waning.
 Accumulation or DistributionLine(A/D):Thismeasures
the volume of trades in a security over time.
accumulation and dispersion across either a short,
medium, or long distance.
 On Balance Volume (OBV): is a metric that measuresthe
volume of securities across time while accounting for
positive and negative flows.
 Relative Strength Index (RSI): These leading signs,
known as the Relative Strength Index (RSI), appear
before trade-related situations.
 DMI (Directional Movement Indicator): A price
indication that contrasts the current share price with
the prior price range is known as a DMI (Directional
Movement indication). The positive figure denotes an
increase in price, whereas the negative value denotes a
decrease in price.
 Movement of Trendlines: Over time, a trendline shows
the uphill and downhill changesinmarketvalues.Create
a collection of practises for using trading logistics to
purchase and sell stocks on the market. Next, ask your
stock broker for API connectivity so you may submit
bids right away. Although it might seem simple, it takes
a lot of time and effort to create a trading bot that can
make millions through high-frequency trading,
something that humans cannot do.
3.2 Stock Market Analysis Using Data Science
Everywhere you look, you can read about the power of data
science. The problem of data affects everyone. Businesses
are interested in learning how data could potentially save
expenses and improve their bottom line. The healthcare
industry is curious about how data science might help them
identify ailments earlier and treat patients more effectively.
Data science is frequently represented by numbers. These
numbers, however, could apply to anything, such as the
number of customers who purchase a product or the
quantity of commodities sold. Of course, these numbers
could also represent money. In order todevelopa distinctive
perspective on the stock market and financial data, data
science is being used in this manner. Stocks, commodities,
and securities all abide by the same basic principles. Over
the past 20 years, trading platforms have become more and
more popular, but each one has its own features, resources,
and fees. Canadians still lack access to zero-commission
trading platforms despite this growing trend. In a 12-month
study, Gary Stevens ofHostingCanada examinedthefeatures
that each of the most well-known stock trading platforms
provides to its users. To make the greatest decision for you,
you must comprehend how they work, and Gary's in-depth
explanation may help you with that. We wholeheartedly
endorse this article from The Balance if you're interested in
learning more about Canadian ETFs.
Recognising Basic Data Science Ideas in the Stock Market In
data science, there are a number of idioms that only
scientists would comprehend.Data scienceisjustmathsplus
a dash of programming and statistics skills.
Algorithm: Data science extensively makes use of
algorithms. A task-completion algorithm is essentially a
series of instructions that must be followed. Probability is
that you are aware of the employment of algorithms in the
buying and selling of stocks. Using algorithms to decide
when to buy or sell stocks is known as algorithmic trading.
For instance, a stock may be programmed into an algorithm
to be bought if it drops by 8% during the course of theday or
sold if it loses 10% of its value since being purchased.
Algorithms are designed to function without the assistance
of humans. They are commonly referred to as bots. They
behave much like robots while making logical decisions.
Training: We're not discussing running a 50-meter dash.
Training is the process of usingdata toeducatea systemhow
to react in machine learning and data science. We are able to
create a learning model. A computer can now make precise
predictions based on historical data thanks to machine
learning. You will need a model of the stock values from the
previous year to serve as a foundation to predict what will
happen if you want to teach a computer to estimate future
stock prices.
Testing: The training set would be the data from January to
October. Then, we'll run tests in November and December.
The predictions made by the computer will be contrasted
with actual prices.
The modeling used to Predict Stock Prices: The
foundation of data science is modelling. With this approach,
past actions are analysed mathematically in order to predict
future outcomes. In the stock market, a time series model is
used. A time series is a collection of data that is indexed
across time, in this case the value of a stock. It is possible to
divide this time span into hours, days, months, or even
minutes. By gathering pricing data using machine learning
and/or deep learning methods, a time series model is
created. Before fitting the model to the data, analysis of the

data is required. This is what makes it possible to anticipate
future stock values over a certain period of time. A second
type of modelling used in machine learning and data science
is a categorization model. These models attempt to
categorise or anticipate what is represented by the data
points when they are provided input. A machine learning
model may be given financial information like the P/E ratio,
total debt, volume, and other factors to determine whether
or not a stock is a good investment when talking about the
stock market or stocks in general.Basedonthefinancial data
we supply, a model can determine if it is the right time to
sell, hold, or buy a stock. It's possible for a model to foresee
something so complex that it ignores the relationship
between the feature and the desired outcome. Overfitting is
the term for this. When a model does not accurately reflect
the data, it is said to be underfit, which leads to overly
straightforward forecasts. If the model has troubledetecting
stock market patterns, overfitting is an issue.When a model
predicts the fundamental average price based on the entire
history of the asset, underfitting occurs. Poor forecasts and
projections are the outcome of both overfitting and
underfitting. We have only begun to explore the connections
between stock market investment and the ideas of machine
learning. But it's important to understand the fundamental
ideas we covered today since they lay the groundwork for
realising how machine learning is used to predict what the
stock market can do. There are more subjects available for
those who wish to go into the details of data scienceandhow
it relates to the stock market.
4. DATA SCIENCE IN STOCK MARKET
Numerous things can be represented by the numbers that
are usually used to represent data. These numbers could be
for sales, inventory, clients, and last but not least, money.
This brings up financial information, specifically the stock
market. Stocks, commodities, securities, and other trading
instruments are largely comparable. Data Science uses a lot
of terms, expressions, and jargon that many people are not
familiar with. We are available to assist with all of that. Data
science requires a working knowledge of mathematics,
statistics, and programming. I'll provide links to numerous
resources throughout the text if you want to read more
about these ideas. Let's go right to what we all wanted to
know now: how to build market evaluations using data
science. To decide whether a stock is worthwhile for
investment, we employ analytics. Now let's go over some
data science concepts that have to do with money and the
stock market. The fields of data science and programming
frequently use algorithms. A set of guidelines that must be
followed in order to finish a specific activity is known as an
algorithm.
A growing trend in the stock market is algorithmic trading,
as you may have heard. Trading algorithms used in
algorithmic trading include parameters like buying a stock
only when it has declined by exactly 5% that day or selling a
stock once it has lost 10% of its value since being bought. All
of these algorithms have the ability to function without the
assistance of a person. Since their trading strategies are
largely mechanical and they trade without emotion,theyare
occasionally referred to as trading bots. Programming,
mathematics, and business are all combined in the
multidisciplinary field of data science and analytics. You
need to understand both words before you can distinguish
between them. Therefore, let's start with data science.The
term "data science" is used to refer to a wide range of
approaches and strategies for obtaining information. In
plainer language. Data science is a field that consists of a
variety of tools, machine learning strategies, and algorithms
for finding patterns in unstructured data. Data analytics is a
process for increasing output and revenue. In this part, data
sets are examined in order to draw conclusions about the
information they contain. Information is extracted,
categorised, and subjecttovariousapproachesdependingon
organisational needs in order todiscoverandassessconduct
information. It was also known as data analysis.
Data Scientist: Computer techniques like neural networks
and machine learning, as well as knowledge of Applied
Statistics, Data Mining, are all necessary. It's vital to be
familiar with database architectures like MySQL, Hive, and
others. Data science is used in a variety of scenarios, such as
online searches and digital advertising. The progress of
machine learning and AI depends on data science. Thendata
analysts construct an algorithm using the data.
Data analysts: The ability to retrieve and query data is
necessary. Data blending, datapurification,data discovering,
and data visualisation are among a data analyst's primary
responsibilities. It is crucial to have a basic understanding of
statistics. The best industries for data extraction by analysts
include healthcare, gaming, and retail.
5. PATTERNS IN STOCK MARKET
Continuation Patterns:An ongoingtrendcanbeconsidered
to come to an end with a continuation pattern. This happens
whenever a bull market pauses during an upswing or a bear
market pauses during a downturn. When a pricing pattern
begins to take shape, it is impossible to tell whether the
trend will last or turn around. To determine if the price
breaks above or below the continuation zone, as well as the
trendlines that were employed to create the price pattern,
pay great attention. Technical analysts frequently predict
that a trend will last as long as it cannot be demonstrated to
have changed.
Reversal patterns: are pricing patterns that showa change
in the direction of the present trend. When the bullsorbears
have peaked, these patterns show it. The existing trend will
halt before continuing in a new direction as fresh energy
emerges from the opposing side (bull or bear).

Pennant: Pennants are continuation patterns that are
created when two trendlines cross. Pennants are
characterized by trendlinesthatmoveinoppositedirections,
one up and one down. Below is an illustration of a pennant.
Frequently, the volume will decrease while the pennant is
built and then increase when the price eventually breaks
through. A bearish pennant pattern denotes a price trend in
the negative direction. Volume is decreasing, and a flagpole
forms on the right side of the pennant, forming a bearish
pattern.
Flag: Flag patterns consist of two parallel trendlines that
may be inclined upward, downward, or horizontally. In
contrast, a flag with a downward bias (bearish) shows as a
break in an uptrending market. A flag with an upward slope
(bullish) appears as a standstill in a downtrending market.
Usually, when a flag form develops, there is a reduction in
volume that is followed by a rise in volume when the price
breaks out of the flag shape.
Wedge: Pennants and wedges are both continuation
patterns made of two trendlines that are convergent;
however, a wedge differs from a pennant in that both
trendlines are moving in the same direction, either up or
down. A wedge with a downwardslopedenotesa break inan
upswing, whereas one withanupwardslopedenotesa pause
in a downturn. while with pennants and flags, volume
frequently declines while patterns form before rising after
the price breaks above or below the wedge pattern.
Ascending Triangle: A trend continuation pattern with a
stated entry point, profit target, and stop loss level is an
ascending triangle. The entrance point is where the
resistance line crosses the breakout line. An ascending
triangle is a bullish trading pattern.
Descending Triangle: A descending upper trend line
suggests that a collapse is about to occur, and a descending
triangle is the opposite of an ascending triangle in that it
suggests declining demand.
Symmetrical Triangles: There is neither an upward nor
downward trend in symmetrical triangles, which emerge
when two trend lines converge in their direction and signal
an impending breakout. As shown in the illustration below,
the size of the breakouts or breakdowns is frequently the
same as the height of the triangle's left vertical side.

Handle and Cup: A bullish continuation pattern called the
cup and handle shows that an upward trend has halted but
will pick up again if the pattern is confirmed. The "cup"
portion of the design should be a "U" shape that resembles
the rounding of a bowl rather than a "V" shape with equal
highs on both sides of the cup.
Head and Shoulders: A reversal pattern known as the head
and shoulders can appear at market peaks or bottoms as a
series of three pushes: an initial peak or trough, a second,
larger one, and a third push that repeats the first.A head and
shoulders top pattern has thepotential toreverseanuptrend
and start a decline. It is almost certain that an upward trend
will resume after a decline that forms a head and shoulders
bottom (or an inverted head and shoulders). As seen in the
image below, trendlines canbe built toconnectthepeaksand
valleys betweentheheadandshoulders.Thesetrendlinescan
be horizontal or slightly slanted. Volume may decreasewhile
the pattern takesshapeand thenriseafterthepricingpattern
is broken.
Double Top and Bottom: These reversal patterns show
areas where the market has twice failed to break through a
support or resistance level. They are used to identify these
areas.A double top is characterised by an initial push-up to a
resistance level, followed by a second unsuccessful attempt,
which results in a trend reversal. It commonly resembles the
letter M.
Gaps: Patterns of reversal are gaps. They appear if a
significant price increase or decrease occurs between two
trading periods. For instance, a stock may close at $5 and
open at $7 after great results or other news. The three
primary categories of gaps are fatigue, runaway, and
breakaway gaps. Breakawaygaps, runaway gaps,andfatigue
gaps all appear during the beginning, middle, and end of
trends, respectively.
6. CONCLUSION
Technical analysis seeks to forecast future changes in
financial price based on past price movements. Because it
doesn't generate precise predictions about the future, think
of technical analysis as being analogous to weather
forecasting. Technical analysis, on the other hand, can help
investors make predictions about what will "likely" happen
to prices over time. Stocks, indexes, commodities,futures, or
any other tradable asset whose price is influenced bysupply
and demand dynamics can all be subject to technical
analysis. Price data is any combinationoftheopen,high,low,
close, volume, or open interest for a certain asset over a

specific period of time (or, as John Murphy prefers, "market
action"). The price information might be daily, weekly, or
monthly as well as intraday (1-minute,5-minute,10-minute,
15-minute, 30-minute, or hourly). Many techniques such as
Machine Learning and Data Science can be used to predict
Stock Market but it can and will never be predicted with
100% accuracy.
REFERENCES
[1] S. Kamil, A. Pinar, D. Gunter, M. Lijewski, L. Oliker, J.
Shalf, "Reconfigurable Hybrid InterconnectionforStatic
and Dynamic Scientific Applications",ACMInternational
Computing Frontiers, 2007.
[2] IPM Homepage: http://ipm-hpc.sourceforge.net/
[3] C. Bell, D. Bonachea, R. Nishtala, K. Yelick,
“Optimizing Bandwidth Limited Problems Using
[4] Robert W. Wisniewski, Peter F. Sweeney, Kartik
Sudeep, Matthias Hauswirth, Evelyn Duesterwald, Calin
Cascavel, and Reza Azimi, "Performance and
Environment Monitoring for Whole-System
Characterization and Optimization", PAC2 (Conference
on Power/Performance interaction with Architecture,
Circuits, and Compilers), 2004.
[5] G. Bell, J. Gray, A. Szalay, “Petascale Computational
Systems,” IEEE Computer, Volume: 39, Issue: 1 Jan.
2006, pp 110-112.
[6] William T.C. Kramer, John Shalf, and Erich
Strohmaier, “The NERSC Sustained SystemPerformance
(SSP) Metric,” LBNL Technical Report LBNL-58868,
September 2005.
[7] S. Kamil, J. Shalf, L. Oliker, and D. Skinner,
“Understanding Ultra-ScaleApplicationCommunication
Requirements,” in Proceedings of the 2005 IEEE
International Symposiumon WorkloadCharacterization
(IISWC), Austin, TX, Oct. 6–8, 2005,pp.178–187.(LBNL-
58059)
[8] Krste Asanovic, et. al., “The Landscape of Parallel
Computing Research: A View from Berkeley, Electrical
Engineering and Computer Sciences,” the University of
California at Berkeley, Technical Report No. UCB/EECS-
2006-183, December 18, 2006. (
http://view.eecs.berkeley.edu/ )

DATA SCIENCE APPROACH TO STOCK MARKET ANALYSIS

More Related Content

Similar to DATA SCIENCE APPROACH TO STOCK MARKET ANALYSIS (20)

More from IRJET Journal (20)

Recently uploaded (20)

DATA SCIENCE APPROACH TO STOCK MARKET ANALYSIS