Continual Traffic Forecasting via Mixture of Experts
This addresses the incremental learning challenge for traffic forecasting in evolving networks, though it is an incremental improvement over existing continual learning methods.
The paper tackles the problem of catastrophic forgetting in traffic forecasting models when new sensors are added over time, proposing a Traffic Forecasting Mixture of Experts (TFMoE) that segments traffic flow into groups with dedicated experts, resulting in superior performance and resilience in experiments on the PEMSD3-Stream dataset.
The real-world traffic networks undergo expansion through the installation of new sensors, implying that the traffic patterns continually evolve over time. Incrementally training a model on the newly added sensors would make the model forget the past knowledge, i.e., catastrophic forgetting, while retraining the model on the entire network to capture these changes is highly inefficient. To address these challenges, we propose a novel Traffic Forecasting Mixture of Experts (TFMoE) for traffic forecasting under evolving networks. The main idea is to segment the traffic flow into multiple homogeneous groups, and assign an expert model responsible for a specific group. This allows each expert model to concentrate on learning and adapting to a specific set of patterns, while minimizing interference between the experts during training, thereby preventing the dilution or replacement of prior knowledge, which is a major cause of catastrophic forgetting. Through extensive experiments on a real-world long-term streaming network dataset, PEMSD3-Stream, we demonstrate the effectiveness and efficiency of TFMoE. Our results showcase superior performance and resilience in the face of catastrophic forgetting, underscoring the effectiveness of our approach in dealing with continual learning for traffic flow forecasting in long-term streaming networks.