This research demonstrates how integrating textual news data with traditional price metrics can enhance stock price prediction accuracy in the Russian securities market.

• Employs LSTM neural networks with a multimodal approach combining numerical and textual data
• Focuses specifically on the Russian market, offering insights for emerging market prediction
• Demonstrates practical engineering applications for financial forecasting systems
• Highlights the importance of news sentiment in price formation beyond traditional metrics

This engineering advancement matters because it offers a more comprehensive approach to financial prediction models, potentially reducing market uncertainty and improving investment decision-making in emerging markets.

Multimodal Stock Price Prediction: A Case Study of the Russian Securities Market