MMDL: A Novel Multi-modal Deep Learning Model for Stock Market Prediction

Stock markets play an important role in modern financial systems. Stock market analysis is of significant interest to both academia and industry. One of the main challenges in this area is modelling multiple sources of information for market forecasting. In this research, we propose MMDL, a Multi-modal Deep Learning model, which utilizes both unique and joint information of the financial news modality and the market data modality for stock predictions. In experiments on real-world datasets, our model clearly outperforms current existing models, which demonstrates that our proposed technique is highly effective for multi-modal stock market predictions.