Automation of stock market management and forecasting using LSTM algorithm

• Published in: AIP conference proceedings Vol. 3075; no. 1
• Main Authors: Vigneshwar, S., Yoganishanth, P., Felix, A. Yovan
• Format: Journal Article Conference Proceeding
• Language: English
• Published: Melville American Institute of Physics 29.07.2024
• Subjects: Algorithms; Decisions; Extreme values; Forecasting; Machine learning; Physical factors; Physiological effects; Predictions; Securities markets
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0223100

One of the most challenging tasks in Computing is predicting the Stock Market. The Stock Market strategies are really Complicated and depend on a large Quantity of data. There are many Components related to forecast, Physical factors, Physiological, rational and irrational, Capitalist Sentiment, market, etc. So, predicting the stock market is always been a challenge for many of them. Many research has been done on predicting stock market using machine learning technique to solve complex problems and improve the prediction values without being explicitly programmed. Stocks with more demand will cost more while stock with large offers will cost less. Because of stock price fluctuations it will affect the investor judgments, this is required to forecast upcoming share prices and stock market details in order to make better information and correct investment decisions. In order to predict future development, the essay suggests the real value bears and the impact of all other occasions. Ideal models and experiences that may be used to create extremely redress expectations can be found using machine learning. LSTM (Long Short-Term Memory) is suggested to show how to look at a stock’s long-term cost. The goal of the research is to forecast stock market prices to generate more precise and familiar investment decisions. An entirely open dataset for stocks is chosen with the following attributes open, high, low, and closing prices and is used to conduct the forecasting.