Sequential Graph Attention Learning for Predicting Dynamic Stock Trends (Student Abstract)

Paper: arXiv 2301.10153

Authors: Tzu-Ya Lai, Wen Jung Cheng, Jun-En Ding

Abstract

The stock market is characterized by a complex relationship between companies and the market. This study combines a sequential graph structure with attention mechanisms to learn global and local information within temporal time. Specifically, our proposed “GAT-AGNN” module compares model performance across multiple industries as well as within single industries. The results show that the proposed framework outperforms the state-of-the-art methods in predicting stock trends across multiple industries on Taiwan Stock datasets.

Complexity vs Empirical Score

• Math Complexity: 5.5/10
• Empirical Rigor: 5.0/10
• Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: This student abstract introduces a graph attention-based architecture (GAT-AGNN) to capture cross-company dependencies and temporal dynamics in the Taiwan stock market. While the integration of graph attention with sequential neural networks is promising and mathematically sound, the extremely short abstract format limits the presentation of comprehensive empirical backtesting, full mathematical derivations, and reproducibility details.

Research Flowchart

  flowchart TD
    A[Research Goal: Predict Dynamic Stock Trends] --> B[Input Data: Taiwan Stock Datasets]
    B --> C[Construct Sequential Graph Structure]
    C --> D[Apply GAT-AGNN Module<br/>Sequential Graph + Attention]
    D --> E[Evaluate Performance:<br/>Multi-Industry vs. Single-Industry]
    E --> F[Outcome: GAT-AGNN Outperforms SOTA Methods]