Stock Market Predictor Web Application

Abstract

This project intends to analyse massive amounts of data and execute thorough regression algorithms on the dataset, which will accurately estimate the future value of a stock using the regression model. The goal of this work is to examine the flaws in the current system and develop a time-series model that would implement more effective algorithms to mostly alleviate them. Anyone may keep track of the favoured stock they wish to invest in using this strategy, and they can maximise their profits by buying plenty of it at the cheapest price. A popular and important topic in academic and financial study is stock price forecasting. Utilizing machine learning, which generates predictions based on the values of current stock market indexes by training on their historical values, a recent development in stock market forecasting technology. Machine learning itself employs a variety of models to facilitate and validate prediction. The study focuses on regression and machine learning for predicting stock value that are LSTM-based.

Introduction

I. INTRODUCTION

The "Stock Market" is discussed in the news each day. You learn about it every time it reaches a new high or low. The rate of investment and business opportunities on the stock market might increase if a trustworthy algorithm could be created to predict the short-term price of a specific stock. Previously, it was possible to predict stock values using artificial neural networks and convolution neural networks, which have an average error loss of 20%. The feasibility of creating a recurrent neural network model that can more precisely predict stock price will be looked at in this study. In addition, if the answer is YES, we will examine how reliable and useful this model is.A stock market is a place where the public can purchase and sell stocks  in companies that are publicly traded..

This model uses the RNN (Recurrent) approach known as Long Short Term Memory (LSTM) and takes into account the historical equity share price of a corporation. The suggested method makes predictions on a specific feature by taking into account the share's historical data that is currently available. Opening price, day High, day Low, previous day o price, close price, date of trading, total trade quantity, and turnover are the characteristics of shares. The suggested model use time series analysis to forecast a share price throughout the specified time period.

The stocks, sometimes referred to as equities, signify ownership in the business. The intermediary that makes it possible to buy and sell shares is the stock exchange. Gaining significant profits is the whole point of making stock price predictions. It's challenging to forecast how the stock market will fare. Other variables, including biological and psychological ones, as well as rational and irrational behavior, are included in the prediction. These forces work together to create a volatile and dynamic market for shares. Because of this, it is quite challenging to create precise stock price predictions.

Stock market return forecasting is a significant problem and a very complex one in financial firms. The task of predicting stock values has always been difficult. The stock prices of any firm have been found to be influenced by both the socioeconomic position of the nation and the company's financial standing. It is no longer closely related to the nation's or specific region's economic progress. Consequently, stock price forecast is now even harder to do than it was previously. Current political events, business-related news, natural disasters, and other factors all have an impact on stock values. One of the key topics that has to be looked into in academic and financial research is stock price prediction.

In conclusion, several organisations in the stock market prediction industry make extensive use of machine learning algorithms.  

A. Problem Statement

Numerous strategies have been put forth in recent days in order to forecast the market price of stocks. To anticipate the future price of a stock market, many sorts of procedures and occasionally combinations of methods have been utilised in these strategies. The feed forward neural network and artificial neural network back propagation algorithm are used in the artificial neural network field of artificial intelligence to anticipate the price of a stock market.

The project's objective is to forecast stock direction and price change using a variety of machine learning algorithms. I utilise regression models to project future prices because the input (Adj Close Price) used in the prediction of stock prices is a continuous value. It is the job of a machine learning or data scientist to analyse the data and develop different algorithms that could be utilised to derive patterns from the historical data of the Microsoft Corporation stock.

B. Parameters used

List of parameters/Symbols used in this paper is listed in

Table 1

C. Research Objective

The aim of this research is to enhance decision-making. The aim of this study is to improve the stock market analysis and forecasting model. As a result, the study's objective is to develop a predictive model for stock prices based on machine learning.

D. Scope of Study

The purpose of our project is to forecast data from the stock market using different algorithms and to evaluate how effectively they perform this task. Both firms and individuals benefit from making well-informed investment decisions. Stock market forecasting is to predict potential changes in a stock's value on a financial exchange. If share price changes can be properly predicted, investors will be able to profit more.

II. RELATED WORK

A. Overview

There were two main indicators for predicting stock prices in the literature. They are fundamental and technical analyses. In the stock market study, both were used. The global economy is significantly impacted by the stock market's current position. People from all different backgrounds, including those in business and academia, have been attracted to the stock market with remarkable success. Due to the nonlinear nature of the stock market, research into it is one of the most significant and well-liked topics worldwide. People use prior research, knowledge, or predictions as the foundation for their stock market investment selections.

Adopting conventional techniques like fundamental and technical analysis doesn't seem to ensure the consistency and accuracy of the predictability. With predictions based on current market values as a result of training on prior values, machine learning technologies have recently become a trend in stock market forecasting. Numerous strategies have been put forth in recent days in order to forecast the market price of stocks. To anticipate the future price of a stock market, many sorts of procedures and occasionally combinations of methods have been utilised in these strategies. The feed forward neural network and artificial neural network back propagation algorithm are used in the artificial neural network field of artificial intelligence to anticipate the price of a stock market.

In [16], the author analyses and projects the stock market index with Markov properties, stock prices, and its interval state in light of the Markov model, giving investors a useful reference model to help them avoid acting irrationally and blindly Building a Markov forecasting model that forecasts the state of an object in the future over a specific length of time using a probability vector for the starting state and a state transition probability matrix is the primary idea behind utilising a Markov chain to make predictions. A stochastic process called a Markov chain has no after-effect characteristics. The after-effect properties imply that the state at time t larger than s only depends on the state in some process at instant s when the state is known at that moment, and does not depend on the state in some process prior to that moment. When a system is balanced, the transition matrix is written as (if the likelihood of the system transitioning from state i to j is Pij) then a transfer matrix is formed by the collection of transition probability vectors in the system state. The author of [17] has developed a brand-new methodology for forecasting stock market prices. For a system that adheres to the Markov chain with unseen states, an HMM is a state machine. The following truths are always true in the HMM with reference to the time series analytic applications if we represent the hidden state at time t as x(t) and the observation at the same time as y(t): X(t) is only reliant on X(t-1), while Y(t) is only dependent on X. (t). Finding the most comparable day in stock market data for a certain day is the first step in the prediction process because it will be used to determine the close value for the following day. To do this, we must first determine the likelihood that prior days fell within the target range. It is simple to calculate the likelihood of a given day from HMM when you have one day's worth of stock data. After computing the probability probabilities for various days, the final step would be to forecast a specific day's close value as the experiment's aim. To do this, they added the probability tolerance parameter, which indicates the neighbourhood of similarity within which we can accept days that are similar to the preceding day. We obtain a list of days that are comparable to yesterday's stock data using the likelihood tolerance, and then we attempt to identify the best estimate as the one with the highest likelihood among all of them.

III. METHODOLOGY

Fundamental, technical, and quantitative technical analysis are three independent schools of thought in trading, all driven by the desire to predict market movements and make money.

A. Fundamental Analysis
Fundamental research includes the analysis of economic factors that influence stock prices. These elements include, for instance, an income statement and a balance sheet. A sort of financial document called a balance sheet lists a company's assets, liabilities, and shareholder equity at a specific point in time. To put it simply, a company's balance sheet informs Intel of its assets, liabilities, and investment costs.

B. Technical analysis      

Technical analysis uses stock volume and price information to forecast what other stockholders will do. Technical analysts use a variety of different types of indicators generated from previous stock price and volume data to predict future stock values.

C. Quantitative Technical Analysis

This distinctive quality of our second school of thinking sets it apart from our later approaches. (Remember that we'll come back to this subject.) Technical Analysis, Quantitative My capstone project's area of interest is stock forecasting. As the name suggests, this kind of stock prediction relies on quantitative prediction techniques rather than machine learning-based graph representations.

There were two important indicators in the literature for stock price forecasting. They are fundamental and technical analysis. Both were used to analyze the stock market

There were two important indicators in the literature for stock price forecasting. They are fundamental and technical analysis. Both were used to analyze the stock market

Stock market prediction is a challenging task to predict the upcoming stock values. It is very difficult to predict because of unstable nature of stock.There were two important indicators in the literature for stock price forecastin. Are fundamental and technical analysis. Both were used to analyze the stock m

Conclusion

The goal of our research study is to help stockbrokers and investors invest money in the stock market. In the stock market, which is a very complex and difficult process because of political difficulties, the financial economic crisis, and many other market-affecting elements, prediction is crucial. Many machine learning algorithms are covered in this survey study, including NLP, Linear Regression, KNN, SVM, LSTM, Artificial Neural Networking, and others. A prediction model\'s value comes from its capacity to guide investors, novices, and shareholders in choosing where to hold or invest their stocks in order to optimise returns while lowering risk. On the other hand, if the dataset contains false information and fabricated news, the stock price. That is untrue. Our long-term objective is to use the new technology to increase the precision of stock prediction. We\'ll develop a model that is more precise than the existing one and get around the limitations at the same time.

References

[1] Sharma, D. Bhuriya, and U. Singh, \"Survey of Stock Market Prediction Using,\" International Conference on Electronics, Communication, and Aerospace Technology, 2017. [2] D. Shah, H. Isah, and F. Zulkernine, \"Stock Market Analysis: A Review and Taxonomy of Prediction Techniques,\" International Journal of Financial Studies, vol. 29, no. 1, pp. 2-22, 2019. [3] S. Bahraini and A. Gheetha, \"Sentiment Analysis for Effective Stock Market Prediction,\" International Journal of Intelligent Engineering and Systems, vol. 14, no. 3, pp. 142-153, 2017. [4] N. N. Reddy, V. K. B P, and N. E., \"Stock Price Prediction Using Sentimental Analysis Using Twitter Data,\" UTC from IEEE Explore, 2020. [5] S. K. Khatri and A. Srivastava, \"Using Sentimental Analysis in Stock Prediction,\" [6] Kaya. M.I.Y., Karsligil. M.E. in 2ND IEEE International Conference on Information and Financial Engineering (ICIFE), pages 478-482, 17th to 19th September 2010. [7] Deju Zhang, Xiaomin Zhang, “Study on Forecasting the Stock Market Trend Based on Stochastic Analysis Method” in International Journal of Business and Management, Vol 4, No. 6, 2009. [8] M. R. Hassan and B. Nath, “Stock market Forecasting using Hidden Markov Models: a new Approach”, Proceedings in 5th IEEE conference on Intelligent Systems Design and Applications, ISDA 2005. [9] Yoon, Y., Guimaraes, T., Swales, G., “Integrating Artificial Neural Networks With Rule- Based Expert Systems”, Decision Support Systems, vol. 11, 1994, pages 497-507, 1994. [10] Yoon Y., Swales G., “Predicting Stock Price Performance: A Neural Network Approach”, Proceedings of the IEEE 24th Annual Conference on System Science, pages 156-162, 1991.

Copyright

Copyright © 2023 Harshita Singh, Sanjana Sahani, Tushar Maddhesiya, Sangam Prasad Prajapati, Sunil Kumar Yadav. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.