An Interpretable Station Delay Prediction Model Based on Graph Community Neural Network and Time-Series Fuzzy Decision Tree

High-speed train delay prediction has always been one of the important research issues in railway dispatching. Accurate and interpretable can enable staff to implement preventive measures scheduling decisions advance, guide relevant departments cooperate completing complex transportation tasks, so as improve rail transit operations, service quality, efficiency operation. This article proposes a new model based on graph community neural network time-series fuzzy decision tree. well capture influence spatiotemporal characteristics, structure, multifactor high-speed station prediction. Besides, time series tree multiobjective optimization reduced error pruning mine potential rules model's interpretability, transparency, high reliability. Finally, we prove that effect proposed is superior than other seven state-of-the-art models our interpretable.