A Distillation-based Future-aware Graph Neural Network for Stock Trend Prediction

ArXiv ID: 2502.10776 “View on arXiv”

Authors: Unknown

Abstract

Stock trend prediction involves forecasting the future price movements by analyzing historical data and various market indicators. With the advancement of machine learning, graph neural networks (GNNs) have been extensively employed in stock prediction due to their powerful capability to capture spatiotemporal dependencies of stocks. However, despite the efforts of various GNN stock predictors to enhance predictive performance, the improvements remain limited, as they focus solely on analyzing historical spatiotemporal dependencies, overlooking the correlation between historical and future patterns. In this study, we propose a novel distillation-based future-aware GNN framework (DishFT-GNN) for stock trend prediction. Specifically, DishFT-GNN trains a teacher model and a student model, iteratively. The teacher model learns to capture the correlation between distribution shifts of historical and future data, which is then utilized as intermediate supervision to guide the student model to learn future-aware spatiotemporal embeddings for accurate prediction. Through extensive experiments on two real-world datasets, we verify the state-of-the-art performance of DishFT-GNN.

Keywords: Graph Neural Networks (GNNs), stock trend prediction, knowledge distillation, spatiotemporal dependencies, Stocks

Complexity vs Empirical Score

• Math Complexity: 7.5/10
• Empirical Rigor: 5.5/10
• Quadrant: Holy Grail
• Why: The paper employs advanced machine learning concepts like graph neural networks, knowledge distillation, and attention mechanisms, supported by mathematical formulations and problem statements. However, its empirical validation is limited to two real-world datasets and lacks detailed backtesting or implementation code, suggesting moderate data and implementation focus.

  flowchart TD
    A["Research Goal:<br>Improve Stock Trend Prediction<br>by capturing future-aware patterns"] --> B["Methodology: DishFT-GNN Framework"]
    B --> C["Teacher Model Learning"]
    C --> D["Student Model Learning"]
    D --> E["Experimentation & Validation"]
    
    subgraph Inputs ["Data & Inputs"]
        F["Real-world Stock Data<br>Spatiotemporal Dependencies"]
    end
    
    F --> C
    
    C --> G["Captures correlation<br>between historical & future<br>distribution shifts"]
    G --> H["Intermediate Supervision"]
    H --> D
    
    D --> I["Future-aware<br>spatiotemporal embeddings"]
    I --> E
    
    E --> J["Key Findings:<br>State-of-the-art Performance<br>on two datasets"]