Utilizing Graph Neural Networks and Game Theory Optimization for Anomaly Detection in Stock Market Investments

Anomaly detection in stock market investments is a critical task for investors and financial institutions to identify irregularities and mitigate risks effectively. This study aims to propose a novel anomaly detection framework by integrating Graph Neural Networks (GNNs) and game theory optimization techniques, leveraging their complementary strengths to enhance anomaly detection accuracy in stock market investments. The proposed framework combines GNNs' capability to capture complex patterns in financial market data with game theory optimization's ability to model strategic interactions among market participants, introducing a novel approach to anomaly detection in stock market investments. The framework utilizes GNNs to construct a graph representation of financial market data, capturing structural dependencies and latent patterns. Game theory optimization techniques are then applied to refine anomaly detection results by modeling strategic interactions among market participants. Evaluation of the proposed framework against existing methods, to demonstrates its effectiveness. The proposed framework achieves a high precision rate of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{8 5. 6 \%}$</tex>, indicating a significant proportion of accurately identified anomalies. The integration of GNNs and game theory optimization in the proposed framework demonstrates promising results for anomaly detection in stock market investments. While maintaining competitive performance metrics, further optimization efforts could enhance its recall while upholding high precision standards, offering valuable insights for investors and financial institutions in navigating the complexities of the stock market.