Evaluation and Scientific Investigation: Stock Market Forecasting Techniques

Ananthi Sheshasaayee; Nivetha S.

Abstract

Analysts and scholars have consistently shown interest in predicting stock market trends, a complex task given the multitude of variables influencing stock values. This article includes a thorough analysis of 50 research papers that propose methodology for stock market prediction, including Bayesian models, fuzzy classifiers, artificial neural networks (ANNs), support vector machines (SVMs) classifiers, neural networks (NNs), and machine learning techniques. The collected papers are categorized using various prediction, clustering methods and variety of datasets. The research gaps and difficulties that the currently used methodologies encounter are listed and explained, assisting the researchers in improving the upcoming efforts. ANN and the fuzzy-based methodology are the most often utilized methods for achieving accurate stock market forecasts. Despite considerable research endeavors, current methods for predicting stock market movements still face notable limitations. Based on the results of this study, it can be inferred that forecasting the future of the stock market is a very difficult process, and various aspects should be taken into account.

Abstract

Keywords