Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY059

Real-Time Stock Forecasting: Leveraging

Live Data and Advanced Algorithms for
Accurate Predictions
Chandu Vaidya1 Gulrukh Nazneen2
Assistant Professor Assistant Professor
S.B. Jain Institute of Technology S.B. Jain Institute of Technology
Nagpur, Maharashtra India, Asia, 441501 Nagpur, Maharashtra India, Asia, 441501
[email protected] [email protected]

Nidhi Singh3 Kapil Katariya4

B.Tech (CSE) B.Tech (CSE)
S.B. Jain Institute of Technology S.B. Jain Institute of Technology
Nagpur, Maharashtra India, Asia, 441501 Nagpur, Maharashtra India, Asia, 441501

Aditya Ramtekkar5 Diptanshu Nasare6

B.Tech (CSE) B.Tech (CSE)
S.B. Jain Institute of Technology S.B. Jain Institute of Technology
Nagpur, Maharashtra India, Asia, 441501 Nagpur, Maharashtra India, Asia, 441501

Diksha Lalmore7
B.Tech (CSE)
S.B. Jain Institute of Technology
Nagpur, Maharashtra India, Asia, 441501

Abstract:- This groundbreaking research introduces an I. INTRODUCTION

innovative stock market prediction methodology that
integrates financial modeling, machine learning, and Analyzing stock market [1] trends is an interdisciplinary
real-time data analysis. Anchored in a deep area of research being undertaken by many disciplines such
understanding of stock market dynamics, including as Finance, Computer Science, Statistics, and Economics.
economic statistics, company performance, and market Stock market is one of the major fields that investors are
sentiment, our approach employs natural language dedicated to, thus stock market price trend prediction is
processing (NLP) and predictive modeling to analyze live always a hot topic for researchers from both financial and
data for accurate stock price predictions. The method technical domains. In this research, our objective is to build a
involves collecting and preprocessing a dynamic dataset state-of-art prediction model for price trend prediction,
enriched with financial indicators and historical stock which focuses on short-term price trend prediction [8]. The
prices. Utilizing Long Short-Term Memory (LSTM) author’s Jingyi Shen, M. Omair Shafiq proposed a fine-tuned
algorithms, our model exhibits an impressive 96% and customized deep learning prediction.
accuracy in forecasting stock movements, showcasing
adaptability to diverse market scenarios and [8] system along with utilization of comprehensive
responsiveness to economic factors and sentiment shifts. feature engineering and combined it with LSTM to perform
The incorporation of live data proves pivotal in prediction. They also applied optimization techniques,
providing timely insights for informed decision-making, such as principal component analysis (PCA) were also
establishing our model as a valuable tool for navigating applied in short-term stock price prediction.
the complexities of the modern financial landscape.
On the other hand, Prof. Bharathi H. N., Kalyani Joshi -
Keywords:- Stock Market, Machine Learning, Natural They used polarity detection algorithm (PCA) [2] for
Language Processing, Prediction. initially labelling news and making the train set. For this
algorithm, dictionary-based approach was used. During the
years, researchers are not only focused on stock price-
related analysis but also tried to analyze stock market trans-
actions such as volume burst risks, which expands the stock

IJISRT24MAY059 www.ijisrt.com 1
Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY059

market analysis research domain broader and indicates this methodologies. Over this period, different way have been
research domain still has high potential. employed to address this challenge.

While the researchers frequently proposed different  Saloni Mullapudi, Sahitya Mullapudi (April 2019) [1] –
neural network solution architectures [8], it brought further Chancing unborn trend for a stock is a pivotal task
discussions about the topic if the high cost of training such because stock trends depend on number of factors. The
models is worth the result or not. Now a days, advanced accuracy of stock price predictions by gathering a large
intelligent techniques based on either technical or amount of time series data and analyzing it in relation to
fundamental analysis are used for predicting stock market. related news articles, using deep learning models. They built
Particularly, for stock market analysis, the data size is huge prediction models based on time seriesv forecasting models,
and also non-linear. To deal with this variety of data such as ARIMA, RNN, and Facebook Prophet So,
efficient model is needed that can identify the hidden completely studied this relationship and concluded that
patterns and complex relations [5] in this large data set. stock trend can be prognosticated using newspapers and
former price history.
This innovative method aims to capitalize on the vast
amount of information available in real-time news sources,  Kalyani Joshi, Prof. Bharathi H.N., [2] –
which can influence market sentiment and subsequently As newspapers prisoner sentiment about the current
impact stock prices. In this cutting-edge application of request, they automate this sentiment discovery and
machine learning to stock market prediction, the algorithm is grounded on the words in the newspapers, can get an overall
trained on a live dataset containing vast previous years data news polarity. The accuracy of the prediction model is 80%
of machine learning models for stock market prediction. and in comparison with news random labelling with 50% of
accuracy; the model has increased the accuracy by 30%.
The aim of the research work is to predict future stock However, also they state that this news impact is good in the
prices based on live data, with the objective of building a request, so more chances of stock price go grandly, If the
predictive LSTM model for forecasting of stocks. news is positive. And if the news is negative, also it may
impact the stock price to go down in trend. used opposition
 Objectives: discovery algorithm for originally labelling news and
making the train set. Random Forest worked very well for all
 Development of a Static Prediction Model: test cases ranging from 88% to 92% accuracy. Accuracy
This involves constructing a system that utilizes LSTM followed by SVM is also considerable around 86%. Naive
(Long Short-Term Memory) algorithms to forecast Bayes algorithm performance is around 83%.
outcomes based on historical data.
 Sai Vikram Kolasani, Rida Assaf, [3]–
 Live Dataset Extraction from Yahoo Finance: In our work, we prognosticate the unborn movement of
The process involves gathering and wrangling real- the United States stock request by assaying the sentiment of
time data from Yahoo Finance, which is then prepared for Twitter posts related to the Stock request. Next, they use the
analysis within the prediction model. sentimental analysis of one year’s data of tweets that
contain the “stock market”, “stocktwits”, “AAPL”
 Interactive Interface Integration: keywords, with the goal of predicting the corresponding
Implementing a user- friendly interface that allows stock prices of Apple Inc. (AAPL) and the US’s Dow Jones
individuals to engage with the prediction model, enabling Industrial Average (DJIA) index prices. We show that
them to select a particular stock to access forecasts in neural networks perform substantially better than traditional
a interactive manner. models for stocks’ price prediction. They got near about
82% of accuracy using SVM model.
II. LITERATURE SURVEY
 Deeksha Chandola, Akshit Meheta [4]–
Stock price trend vaticination is an active exploration The paper presents a deep literacy model that helps the
area, as more accurate prognostications are directly related investors comprehend the request’s trading geste This
to further returns in stocks. thus, in recent times, significant model outperforms previously proposed models that have
sweats have been put in into developing models that can used news of past weeks and months, with improved
prognosticate for unborn trend of a specific stock or overall accuracy of 65.4%. The frame combines word embedding
request. utmost of the being ways make use of the with intermittent neural network for. The result supports the
specialized pointers. Some of the experimenters showed that hypothesis that the information in the news headlines has a
there's a strong relationship between news composition about short temporal effect on the investors prognosticating stock
a company and its stock prices oscillations. Following is price directional movement. The model takes a combination
discussion on former exploration on sentiment analysis of of fiscal time series and news captions as input. Compared to
textbook data and different bracket ways. This exploration other applicable work, the use of mongrel input
paper makes a significant donation background and appreciatively influences the affair. Regarding the textbook
literature review. The assessment of private responses has representation, it's observed that both good news and bad
been a longstanding pursuit, with nearly two decades of news induce a change in the stock price. It helps in the
exploration devoted to the development and refinement of macroeconomic analysis of the stock. Hence, the news titles

IJISRT24MAY059 www.ijisrt.com 2
Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY059

from the day ahead are aligned in a unique case and used for This approach helps lay the groundwork for a
directional soothsaying. fundamental comprehension of market behavior. Stock price
prediction is significantly aided by machine learning models.
 Mehar Vijh, Deeksha Chandola, (Noida 201304, India) Understanding the sentiment of the market can be gained
[5]– through sentiment analysis of news and social media. Textual
In this work, Artificial neural Network and Random data can be analyzed to determine the general sentiment of
forest techniques have been utilized for predicting the next market participants with the aid of Natural Language
day closing price for five companies belonging to different Processing (NLP) tools.
sectors of operation. The historical dataset available on
company’s website consists of only few features like high, In the initial phase of our research, we commenced with
low, open, close, adjacent close value of stock prices, the acquisition of raw data from live sources encompassing
volume of shares traded etc.,which are not sufficient enough. stock prices, trading volumes, economic indicators, and
Results show that the best values obtained by ANN model sentiment scores derived from news feeds. Following this, a
gives RMSE (0.42), MAPE (0.77) and MBE (0.013). meticulous process of feature extraction was undertaken to
distill pertinent information, including stock prices, trading
[9] Here authors have survey the portfolio optimization volumes, and sentiment scores, establishing a
technique, A few methods and approaches to make it comprehensive feature set for subsequent analysis. The
simpler and more effective are the following: Market Cap identification of original indices and potential expansion of
Weighted Portfolio; Sharpe Ratio; Mean-Variance features ensued, facilitating a nuanced understanding of the
(Markowitz); Minimum Variance; Hierarchical Risk Parity dataset's structural composition. Employing Recursive
(HRP); Cvar and Var; Equal Weighted Portfolio; and Feature Elimination (RFE) enabled the systematic selection
Minimum Variance. This study paper provides a detailed of the most impactful features, while particular emphasis
overview of all of these methods and approaches. was placed on those exhibiting high weights or importance
scores. To further refine the feature set, Principal
 Findings Component Analysis (PCA) was employed, serving as a
Initially, we utilized the LSTM algorithm, achieving an pivotal step in dimensionality reduction while retaining
impressive 96% accuracy in training our stock market essential variance within the high-weighted features. This
prediction model. Subsequently, we developed a user multistep process laid the foundation for subsequent stages.
interface featuring two graphs: a time series graph providing
accurate predictions until the current date and a forecasting  Dataset:
graph offering precise predictions for the next 2 years, with a In this segment, we delve into the intricacies of
provided range for the subsequent 3 years. This professional our data extraction journey from various public sources,
approach enhances user accessibility and transparency in unveiling the meticulous crafting of our final dataset. Given
communicating the model's capabilities and limitations, the inherent diversity of stock market-related data, our
positioning our research at the forefront of stock market approach commenced with a rigorous comparison of
prediction methodologies. existing works within the realm of financial research.

 Proposed Approach Once our data arsenal was assembled, we undertook the
The core of our approach lies in the development of a crucial task of defining a robust data structure for the
sophisticated machine learning model, incorporating dataset. This structural framework served as the backbone for
algorithms like Long Short-Term Memory (LSTM). This organizing and interpreting the wealth of information at our
model is trained on historical data with feature engineering disposal. Presented below is an exhaustive exploration of
to discern critical variables influencing stock prices. our dataset, offering a nuanced understanding of its
Additionally, sentiment analysis on news and social media architecture, data tables within each category, and a detailed
data, facilitated by Natural Language Processing (NLP) segment definition. This holistic exposition provides a
tools, is integrated to gauge market sentiment—a crucial unique insight into the intricacies of our data curation
factor in stock price movements. process and lays the foundation for the subsequent analyses
and findings presented in this research endeavor.
First and foremost, gathering real-time data is essential.
Financial data, such as stock prices, trade volumes,  Description of Dataset:
economic indicators, and news mood, can be retrieved in real- In this research endeavor, we employed the yfinance
time for this purpose. For this, news organizations, market Python library as a pivotal tool for acquiring our financial
data providers, and financial platforms' APIs are frequently dataset. Yfinance facilitated the seamless extraction of
used. diverse financial data directly from Yahoo Finance. This
dynamic Python library stands out for its simplicity and
Statistical analysis can then be used to find patterns convenience, offering an accessible means to access an
and trends within the data. While correlation analysis can extensive array of financial metrics. The dataset
show connections between stock prices and different encompasses essential components such as historical stock
economic factors, time series analysis aids in understanding prices, trading volumes, and other pertinent indicators for a
past price movements. This examination. multitude of companies.

IJISRT24MAY059 www.ijisrt.com 3
Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY059

By leveraging yfinance, we ensured a robust and up-to- enhancing our analysis of the intricate relationship between
date dataset, providing a foundation for our predictive market sentiment and stock price movements. This
model. The incorporation of live news feed data further meticulously curated dataset, drawn from yfinance.
enriched the dataset, introducing real-time insights and

III. METHODOLOGY

 Following System Design (Figure 1) is Proposed in this Project to Classify News Articles for Generating Stock Trend Signal.

Fig 1 Flowchart of Propose System

Methodology is mainly divided into 3 phases. In ratio of 60-40 to ensure effective model evaluation. Prior to
Phase 1 we started our project with identifying problem in training, we diligently preprocessed the raw data to render it
stock market where we analyse that stock market prediction suitable for the construction and training of our LSTM
is mainly depend on the previous pattern or trend of the model.
particular stock. We gathered requirement for making a
Deep Learning Model which can predict the stock. We Subsequently, the model underwent a rigorous training
decide which technology to be used and proposed our work. process, during which it learned patterns and relationships
Result of phase 1 is to Developed such model which can inherent in the static stock dataset. This training phase
predict the stock by taking live data. In phase 2, We collect resulted in the attainment of desired outputs, which were
the live stock data by implementing a library named yfinance further validated on the testing dataset. The outcomes from
which take data from yahoo finance site then we processed the testing phase demonstrated that our model effectively met
that data to make used of data in our model. We used LSTM the specified requirements, affirming the successful
Algorithm to create our model. At the end of phase 2, we accomplishment of our primary objective: the construction of
have developed the model which can predict the stock. In a well-performing model using a static stock dataset.
phase 3, We started making our User Interface in streamlit
framework of python Then we implement our model in As we move forward, the subsequent stages of our study
python pycharm to connect them. As a result of phase 3 , we will involve the introduction of dynamic elements, such as
get our desired output which is prediction of stock. live news feed data, to enhance the predictive capabilities of
our model in the ever-changing landscape of the stock
IV. IMPLEMENTATION market.

 Model Building:  Providing Live Data to Our Model:

In the initial phase of our study, we constructed a Our second objective focuses on equipping our model
robust model employing the LSTM (Long Short-Term with the capability to predict stocks in real-time by
Memory) algorithm using a static dataset. The dataset was incorporating live data. To achieve this, we utilized the
meticulously divided into training and testing sets, with a yfinance Python library, a powerful tool that facilitates the

IJISRT24MAY059 www.ijisrt.com 4
Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY059

seamless retrieval of financial data from Yahoo Finance. prediction. The system then provides search results,
This library offers a user- friendly and efficient means of presenting the predicted stock outcomes. Our ultimate goal
accessing a diverse range of financial information. with this implementation is to deliver an accessible and
interactive interface that empowers users to seamlessly
In the process, we integrated our model with the interact with and interpret stock predictions, fostering a user-
yfinance library, enabling it to dynamically fetch live data friendly and informative experience.
directly from the Yahoo Finance website. This integration
transforms our model into a real-time application, providing Hence, our model has evolved into a powerful tool for
timely and up-to- date information for stock market stock prediction, seamlessly blending the prowess of
predictions. The development environment for this real-time algorithms like Long Short-Term Memory (LSTM), the
model is PyCharm, a versatile Python IDE, where the versatility of libraries such as yfinance for live data
integration with yfinance takes place seamlessly. integration, and the user-friendly interface afforded by the
Streamlit framework. With these components harmoniously
As we progress, the utilization of live data ensures that integrated, our model stands prepared to make real-time
our model remains adaptive and responsive to the dynamic stock predictions while engaging users through an intuitive
nature of financial markets, enhancing its practical utility for interface. By marrying cutting-edge technology with user-
real- world stock market predictions. centric design, our comprehensive solution embodies the
intersection of efficiency, accuracy, and accessibility in the
 User Interface: realm of stock market prediction.
In the final stage of our implementation, we aimed to
furnish our model with a user-friendly interface, allowing V. RESULTS
users to effortlessly access stock predictions. To materialize
this objective, we leveraged the Streamlit framework—an Our findings showcase the model's adaptability to
open- source tool designed for the rapid development and various market scenarios, emphasizing its robustness in
sharing of aesthetically pleasing machine learning and data capturing trends and responding to shifts in economic
science web applications. factors and sentiment. The incorporation of live data,
including financial indicators and sentiment analysis,
The integration of Streamlit proved instrumental in significantly contributes to the model's predictive
simplifying the implementation of our user interface. The capabilities, providing investors with timely insights for
framework's user-friendly features facilitated the creation of informed decision-making. These results affirm the potential
an interactive interface, streamlining the user experience. of our approach in advancing the field of stock market
This interface enables users to input the desired stock- prediction, offering a valuable tool for navigating the
predicting company name and the number of years for complexities of today's dynamic financial landscape.

Fig 2 Time Series Graph

Here, figure 2 graph encapsulates stock predictions up today, it provides a comprehensive visualization of predicted
to the current date, employing data sourced from Yahoo stock trajectories.
Finance. This time series graph serves as a reliable indicator,
offering accurate predictions for diverse companies. By
forecasting the performance of various entities leading up to

IJISRT24MAY059 www.ijisrt.com 5
Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY059

The utilization of data extracted from Yahoo Finance and strategic decisions. Investors can leverage this resource
enriches the graph with dynamic and up-to-date information, to comprehend stock market movements and optimize their
enabling investors to gain valuable insights into predicted investment strategies based on the latest forecasts. In
trends. This real-time data integration transforms the graph essence, the graph serves as a valuable ally for navigating
into a practical tool, assisting investors in making informed the complexities of the stock market landscape.

Fig 3 Forecasting Graph

The figure 3 showcased above paints a forward- Rida Assaf, the authors explore the dynamic realm of stock
looking picture, projecting trends for multiple companies market prediction through sentiment analysis derived from
over the next five years. Driven by data meticulously Twitter feeds. Notably, their methodology leverages the
extracted from Yahoo Finance, this tool emerges as a Support Vector Machine (SVM) algorithm to forecast future
valuable resource for gaining insights into the anticipated stock movements, achieving a commendable accuracy rate
trajectories of various stocks. of 83%. We focus on the utilization of live data for
prediction, providing a real-time perspective on stock trends.
Notably, the graph provides precise results for the Moreover, our approach employs the LSTM algorithm,
immediate next 1-2 years, offering a clear snapshot of resulting in a remarkable accuracy of 96% during the
expected stock movements. Additionally, for the subsequent training phase.
3-4 years, the tool furnishes a range within which the stock
is predicted to lie. This dynamic feature enhances flexibility In the research paper titled "Stock Trend Prediction
and accommodates varying investment strategies. Using News Sentiment Analysis," authored by Kalyani
Joshi, Prof. Bharathi H. N., and Prof. Jyothi Rao, the authors
Users are empowered with the ability to customize explore the realm of predicting current stock trends through
their predictions by selecting the number of years for sentiment analysis of news articles. Employing the Random
projection, tailoring the tool to meet their specific Forest algorithm, their model attains a substantial accuracy
forecasting needs. In essence, this forecasting graph emerges of 90%. In contrast, our research takes a unique trajectory,
as a forward-thinking companion, aiding investors in concentrating on the prediction of future stock movements
navigating the intricacies of the financial landscape with using live data.
foresight and strategic acumen.
In the research paper titled "Stock Price Prediction
VI. DISCUSSION Using News Sentiment Analysis," authored by Saloni
Mohan1, Sahitya Mullapudi1, Sudheer Sammeta1, Parag
In this section, we discuss and compare the results of Vijayvergia1 and David C., the authors adopted a distinct
our proposed model, other approaches, and the most related approach by utilizing Recurrent Neural Networks (RNN) for
works. their predictive model. RNNs are well-suited for capturing
sequential dependencies in data, making them particularly
 Comparison with related works relevant for time-series prediction tasks such as stock price
forecasting. This choice of algorithm aligns with the
 (SVM), Random Forest (RF), RNN and Convolutional objective of incorporating the temporal dynamics of news
Neural Network (CNN). sentiment to enhance the accuracy of stock price predictions.
In the research paper titled "Predicting Stock It's worth noting that our research, in contrast, employs the
Movement Using Sentiment Analysis of Twitter Feed with Long Short- Term Memory (LSTM) algorithm for a similar
Neural Networks," authored by Sai Vikram Kolasani and purpose, showcasing the diverse range of neural network

IJISRT24MAY059 www.ijisrt.com 6
Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY059

architectures explored within the domain of stock market strategy aligns with contemporary cloud computing
prediction. paradigms, making it a relevant and forward-thinking
solution for modern data management challenges. the
In the research paper titled "Forecasting Directional multifaceted advantages of the device- independent
Movement of Stock Prices using Deep Learning," authored replication approach extend beyond immediate concerns of
by Deeksha Chandola, Akshit Mehta, and Shikha Singh, the data accessibility and loss prevention. It addresses critical
authors conducted a comparative analysis of the Stock aspects of disaster recovery, resource optimization, and
Movement Prediction (SMP) model using Recurrent Neural scalability, making it a comprehensive and forward-looking
Networks (RNN) and Long Short-Term Memory (LSTM) solution in the ever-evolving field of data management
architectures. Notably, their study focused on forecasting the within cloud computing environments.
directional movement of stock prices. However, it's
important to highlight that the data utilized in their analysis FUTURE SCOPE
did not involve live data. Despite this, their findings
demonstrated an accuracy of 65% in predicting the The future scope of "Stock Market Prediction Using
directional movement of stock prices. This study contributes Live Data" is vast and dynamic. By embracing technological
valuable insights into the application of deep learning advancements, expanding the scope to global markets, and
techniques for stock price forecasting, offering a fostering collaboration, the project has the potential to
comparative evaluation of RNN and LSTM models. revolutionize the way investors navigate the complex
landscape of financial markets. Stock market prediction can
As we delve into the comparison with related work, it be use in the areas such as: Incorporation of Advanced NLP
becomes evident that our research not only builds upon Techniques, Real-time Market Impact Assessment,
existing knowledge but introduces a sophisticated paradigm Continuous Model Optimization with Reinforcement
that enhances predictive capabilities in the dynamic realm of Learning, Collaboration with Financial Institutions.
stock market forecasting. From the previous works, we
found the most commonly exploited models for short-term The future scope of stock market prediction using live
stock market price trend prediction are support vector data on live datasets holds immense potential for
machine. revolutionizing investment strategies and financial decision-
making. Incorporating real-time news data into stock market
VII. CONCLUSION prediction models can provide a more dynamic and accurate
assessment of market trends, enabling investors to make
The conclusion of the study highlights the effectiveness informed decisions. With the advent of advanced machine
of the proposed approach, which involves the generation of learning and natural language processing techniques, the
two copies of data, demonstrating a device-independent analysis of news sentiment, market sentiment, and other
replication method. This innovative strategy ensures that relevant factors can be automated, offering investors a
data remains accessible even if deleted from one device, as comprehensive understanding of the factors influencing
it is retained on another. The significance of this approach stock prices.
lies in its potential to instill trust among users, enhance user
satisfaction, and mitigate data loss risks. The study Furthermore, the integration of live datasets ensures
underscores the importance of data replication as a potent that the predictive models to changing market conditions in
technique to minimize user waiting times and augment data real time. This responsiveness is crucial in the ever-evolving
availability. By efficiently distributing workload across landscape of financial markets where news and events can
cloud nodes, the proposed approach contributes to improved have an instantaneous impact. The application of deep
response times, addressing the crucial aspect of user learning algorithms, coupled with the continuous influx of
experience. The conclusion emphasizes the necessity of a live data, can enhance the predictive accuracy of these
replica replacement method, particularly in the context of models, making them more reliable for investors seeking to
restricted storage space, to optimize the efficiency of navigate the complexities of the stock market. Additionally,
dynamic replica management. Overall, the findings suggest advancements in technology, such as improved data
that the outlined approach not only introduces a resilient and processing speeds and enhanced computing capabilities, will
reliable data replication system but also acknowledges the contribute to the scalability of these predictive models. This
practical challenges associated with storage limitations, scalability allows for the analysis of vast amounts of live
offering a comprehensive solution for enhancing overall data, leading to more robust and nuanced predictions. As the
system performance and user satisfaction. field of artificial intelligence continues to evolve, so too will
the sophistication of stock market prediction models,
Furthermore, the conclusion underscores the role of the potentially unlocking new insights and uncovering
proposed approach in optimizing resource utilization within previously unnoticed patterns within the data.
cloud computing environments. By efficiently distributing
workload across cloud nodes, the approach not only
enhances response times but also minimizes the risk of
resource bottlenecks. This is particularly crucial in dynamic
and scalable computing environments where the demand for
resources can vary rapidly. The study emphasizes how the

IJISRT24MAY059 www.ijisrt.com 7
Volume 9, Issue 5, May – 2024 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165 https://doi.org/10.38124/ijisrt/IJISRT24MAY059

REFERENCES [14]. Singh, Tinku & Kalra, Riya & Mishra, Suryanshi &
Singh, Satakshi & Kumar, Manish. (2022). An
[1]. Saloni Mohan, Sahitya Mullapudi, “Stock Price efficient real-time stock prediction exploiting
Prediction Using News Sentiment Analysis”, IEEE incremental learning and deep learning. Evolving
Fifth International Conference on Big Data Systems. 14. 10.1007/s12530-022- 09481-x.
Computing Service and Applications [15]. Ho, Kin-Yip & Wang, Wanbin. (2016). Predicting
(BigDataService) April 2019. Stock Price Movements with News Sentiment: An
[2]. Prof. Prof. Bharathi H. N., Kalyani Joshi, “Stock Artificial Neural Network Approach. 10.1007/978-3-
Trend Prediction Using News Sentiment Analysis”, 319-28495- 8_18.
International Journal of Computer Science & [16]. Kim T, Kim HY. Forecasting stock prices with a
Information Technology (IJCSIT), Department of feature fusion lstm-cnn model using different
Computer Engineering, June 2016, KJSCE, Mumbai. representations of the same data. PloS one.
[3]. Kolasani, S.V. and Assaf, R. (2020) “Predicting 2019;14(2):0212320. doi:
Stock Movement Using Sentiment Analysis of 10.1371/journal.pone.0212320.
Twitter Feed with Neural Networks”, Journal of Data
Analysis and Information Processing, 8, 309-319.
[4]. Chandola, D., Mehta, A., Singh, S. et al. “Forecasting
Directional Movement of Stock Prices using Deep
Learning”. Ann. Data. Sci. 10, 1361–1378 (2023).
[5]. Mehar Vijh, Deeksha Chandola, Arun Kumar, “Stock
Closing Price Prediction using Machine Learning
Techniques”, International Conference on
Computational Intelligence and Data Science
(ICCIDS 2019), 167, 599–606 (2020).
[6]. Jagruti Hota, Sujata Chakravarty, Stock Market
Prediction Using Machine Learning Techniques,
Faculty of Management Studies, Sri University,
Odisha, India.
[7]. Jianxin Bi, “Stock Market Prediction Based on
Financial News Text Mining and Investor Sentiment
Recognition”, Mathematical Problems in
Engineering, vol. 2022, Article ID 2427389, 2022.
[8]. Shen, J., Shafiq, M.O. “Short-term stock market price
trend prediction using a comprehensive deep learning
system”. J Big Data 7, 66 (2020).
[9]. Vaidya, C., Poharkar, S. S., Tandon, M. H. K.,
Jaulkar, P. S., Dalvi, S., Singh, A., & Bhure, K.
(2023, April). Concrete survey and analysis on
portfolio optimization techniques. In AIP Conference
Proceedings (Vol. 2753, No. 1). AIP Publishing.
[10]. Raut Sushrut Deepak, Shinde Isha Uday, Dr. D.
Malathi, “Machine learning approach in stock market
prediction”, International Journal of Pure and
Applied Mathematics, Special Issue, 1311-8080, 71-
77 (2017).
[11]. Gajamannage, Kelum & Park, Yonggi, “Real-time
Forecasting of Time Series in Financial Markets
Using Sequentially Trained Many-to-one LSTMs”
(2022).
[12]. Wen, Min & Li, Ping & Zhang, Lingfei & Chen,
Yan. (2019). Stock Market Trend Prediction Using
High-Order Information of Time Series. IEEE
Access. PP. 1-1. 10.1109/ACCESS.2019.2901842.
[13]. Dr YVS Sai Pragathi, M V S Phani Narasimham and
Dr B V Ramana Murthy, Analysis and
implementation of realtime stock prediction using
reinforcement frameworks, CSE Dept, Stanley
College of Engineering & Technology for Women,
Hyderabad, India

IJISRT24MAY059 www.ijisrt.com 8