Olympic Medal Prediction via Adaptive Triple-Fusion: Combining AAC Model with ARIMA-State Space Dynamics

The Olympic Games represent the most influential sporting event globally. It is of great importance to predict the distribution of medals for its strategic meaning in national sports programs. However, the existing prediction methods often fail to capture both macro and micro factors. To improve prediction accuracy and explainability, we propose a model that combines the AAC (Senior Athletes-New Athletes-Country) framework with ARIMA-state space fusion methodology. This model integrates athlete development with national trends using hierarchical Bayesian networks and balances short-term fluctuations with long-term tendencies through an adaptive integration mechanism. When applied to the 2024 Olympics, the model achieved an absolute average prediction error of 4.1% for total medals and 3.9% for gold medals. These results provide valuable new insights and tools for enhancing Olympic prediction and sports event forecasting systems.