Stock Price Forecasting Using Machine Learning Techniques: A Systematic Review

Investment strategies in the stock market are intricate, hinging upon the analysis of extensive datasets. Recently, there has been a growing interest in leveraging machine learning techniques to determine their potential in enhancing market predictions compared to conventional methods. This study aims to pinpoint potential pathways for future research in machine learning-driven stock market forecasting by delving into existing literature. Employing a systematic literature review approach, relevant peer-reviewed journal articles from the past two decades are identified and grouped based on their shared methodologies and contexts. From this analysis, four key categories emerge: investigations employing artificial neural networks, research centered on support vector machines, inquiries that fuse genetic algorithms with other approaches, and those integrating hybrid or alternative artificial intelligence methodologies. Each category's studies are scrutinized to uncover shared discoveries, distinctive insights, limitations, and areas demanding further exploration. The conclusive segment presents overarching deductions and outlines for prospective research endeavors.