Predicting Stock Market Price Using Machine Learning Techniques

Abstract

Financial time-series predictions like stock and stock indexes have become the main focus of research because of their fluctuating and nonlinear nature in almost all advanced and developing countries. Predicting stock market prices is a crucial topic in the present economy as multiple factors like the global economy, political conditions, country’s performance, company’s financial reports, and many more affect the stock price. Hence, the inclination toward new opportunities to predict the stock market has increased dramatically among professionals. Thus, many predictive techniques are employed over the past few years to maximize the profit and diminish the losses from the stock market movements. With the advancement of artificial intelligence and increased computational capabilities, various methods with programming models have been proven to be more efficient in forecasting stock trends. Mostly, the data size in the stock market is huge and not linear. So, efficient models are required to deal with the complexity and nonlinearity of huge datasets and to find out the hidden pieces of information. Therefore, an effort has been made to forecast the future stock market prices by applying various machine learning techniques such as linear regression (LR), support vector machine (SVM), decision tree (DT), and long short-term memory (LSTM). Then the performance parameters of all ML models such as the root mean squared error, mean absolute error, and mean square error are computed. Our experimental results show that LSTM provides better accuracy in terms of forecasting stock prices compared to the SVM and decision tree algorithm.