Related Experiment Videos
Intelligent financial forecasting using transformers, neuro-symbolic AI, and agent-based systems
V Jeyajeev1, R Jagadeesh Kannan1, R Deebalakshmi1
1SRM Institute of Science and Technology Tiruchirappalli, Tiruchirappalli, 621105, Tamil Nadu, India.
Scientific Reports
|May 30, 2026
Summary
This study introduces an AI framework using a transformer model and LLM-based decision-making for accurate stock price prediction. It enhances financial forecasting with interpretable and adaptable AI-driven strategies.
Area of Science:
- Artificial Intelligence
- Financial Forecasting
- Deep Learning
Background:
- Stock market forecasting is challenging due to volatile prices and complex temporal dependencies.
- Existing models struggle with unstable trends, leading to inaccurate predictions.
- Need for adaptable and interpretable AI-driven financial forecasting models.
Purpose of the Study:
- To develop an AI-driven framework for accurate and understandable stock price prediction.
- To integrate a sequence-to-prediction transformer with LLM-based decision-making.
- To enhance financial forecasting accuracy and interpretability.
Main Methods:
- Utilized a transformer-enabled deep learning approach for price forecasting.
- Incorporated multi-head attention mechanisms for trend identification.
- Employed Neuro-Symbolic AI (NSAI) and Advanced AI Agent Architecture for decision-making and validation.
Main Results:
- The AI framework demonstrates improved adaptability to market changes.
- Combined deep learning with symbolic reasoning for accurate and interpretable predictions.
- The Advanced AI Agent Architecture enhanced decision quality using LLM-based information.
Conclusions:
- The proposed AI framework offers a flexible and explainable approach to stock market forecasting.
- This research advances financial market analysis through interpretable AI strategies.
- The findings can lead to improved investment methods for traders and investors.
