Deep Learning in Financial Markets: A Systematic Literature Review of Methods and Future Direction for Price Prediction

This systematic literature review (SLR) investigates the application of deep learning methods in financial market price prediction, focusing on the most studied financial markets, commonly used deep learning techniques, input variables, performance evaluation metrics, and proposed future research directions. The review covers 56 peer-reviewed articles published between 2020 and 2024, selected from ACM Digital Library, IEEE Xplore, and ScienceDirect databases, examining the stock, cryptocurrency, commodity, and forex market price prediction. Long Short-Term Memory (LSTM) and Gated Recurrent Units (GRU) are identified as the most widely used methods due to their capability in handling long-term dependencies in time series data. The review highlights future directions in hyperparameter optimization, hybrid model integration, and the use of social network data. This study fills the gap in literature by analyzing recent advancements and their applicability across various financial markets.