International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 03 | Mar 2025
p-ISSN: 2395-0072
www.irjet.net
A Comparative study of Stock Market Investment using Long ShortTerm Memory J Gowri Thangam 1, M Maanis2 and M Manasa3 1Assistant Professor, Department of Computer Applications, PSG College of Technology, Coimbatore, India 2UG Student, Department of Economics, PSG College of Arts and Science, Coimbatore, India 3UG Student, Department of Statistics, PSG College of Arts and Science, Coimbatore, India
---------------------------------------------------------------------***--------------------------------------------------------------------The complexity of stock market prediction arises from its Abstract - Stock market prediction is a crucial field of
dynamic and volatile nature, influenced by numerous factors such as economic indicators, company performance, geopolitical events, market sentiment, and technological advancements. Techniques for prediction range from traditional methods like fundamental and technical analysis to modern approaches leveraging machine learning, artificial intelligence, and data analytics.
study that impacts financial data to predict prices of stocks and trends, aiding financial organizations and investors in decision-making. This research explores advanced predictive modeling techniques by integrating real-time market sentiment, macroeconomic variables, and historical stock data. The deep learning approach including regression-based modeling, neural networks with recurrent features (RNNs), and long short-term memory (LSTM) models are used to identify challenging patterns and non-linearity in the data set. The study highlights the significance of data pre-processing, feature selection, and evaluation measures including mean squared error (MSE) and accuracy. The preliminary results indicate that the MSE has diminished, demonstrating that deep learning algorithms may enhance stock market forecasts. The realtime prediction models can be improved by combining adaptive algorithms and reinforcement learning which will pave pathway for future research.
Despite advancements in predictive models, uncertainty remains a fundamental challenge in the stock market. Understanding the strengths and limitations of different methods and integrating them effectively is key to enhancing the reliability of market predictions and achieving financial goals.
2. Related Works The financial and economic importance of stock market prediction has attracted a lot of attention and research. Numerous approaches have been put forth and put into practice, ranging from sophisticated machine learning algorithms to statistical models. This literature review provides an overview of the key theories and findings related to stock market prediction (Das, 2024)..
Key Words: Deep Learning, Forecasting, Investors, LSTM, MSE, RNN, Stock Market.
1. INTRODUCTION Stock market prediction is the process of forecasting future stock prices or market trends using various analytical methods, models, and tools. It is a critical area in finance, as accurate predictions can guide investors in making informed decisions, maximizing returns, and minimizing risks. Predicting stock market movements involves estimating future changes in stock prices or market trends using a variety of models, strategies, and algorithms. By examining technical indicators, historical data, and other relevant aspects, stock market prediction aims to forecast the direction of stock indices or prices. In order to maximize profits or minimize losses, these projections assist traders, analysts, and investors in making well-informed decisions regarding the purchase, sale, or holding of stocks (Lin, 2022). Because of the stock market's intrinsic complexity and wide range of influencing factors, predicting is both difficult and essential. While bad forecasts can cause large financial losses, accurate forecasts can result in lucrative investments.
© 2025, IRJET
|
Impact Factor value: 8.315
2.1. Statistical Approaches Statistical approaches include historical market data, statistical models, and mathematical techniques to predict future stock prices, trends, or market changes in stock market forecasting. These methods seek to recognize patterns, relationships, and trends that can be used to forecast future behavior in financial markets. 2.1.1 Autoregressive Integrated Moving Average (ARIMA) The Autoregressive Integrated Moving Average (ARIMA) is a popular tool for time series forecasting because of the extent to which it models linear trends and how easy it is to use. According to studies, ARIMA does well for shortterm forecasts but has trouble with complex and nonlinear financial trends. (Box et al., 2015).
|
ISO 9001:2008 Certified Journal
|
Page 510