Abstract

Metro Origin-Destination (OD) prediction is a crucial yet challenging spatial-temporal prediction task in urban computing, which aims to accurately forecast cross-station ridership for optimizing metro scheduling and enhancing overall transport efficiency. Analyzing fine-grained and comprehensive relations among stations effectively is imperative for metro OD prediction. However, existing metro OD models either mix information from multiple OD pairs from the station's perspective or exclusively focus on a subset of OD pairs. These approaches may overlook fine-grained relations among OD pairs, leading to difficulties in predicting potential anomalous conditions. To address these challenges, we learn traffic evolution from the perspective of all OD pairs and propose a fine-grained spatialtemporal MLP architecture for metro OD prediction, namely ODMixer. Specifically, our ODMixer has double-branch structure and involves the Channel Mixer, the Multi-view Mixer, and the Bidirectional Trend Learner. The Channel Mixer aims to capture short-term temporal relations among OD pairs, the Multi-view Mixer concentrates on capturing spatial relations from both origin and destination perspectives. To model long-term temporal relations, we introduce the Bidirectional Trend Learner. Extensive experiments on two large-scale metro OD prediction datasets HZMOD and SHMO demonstrate the advantages of our ODMixer. Our code is available at https://github.com/KLatitude/ODMixer

Framework

Experiment

Conclusion

In this paper, we introduce ODMixer, a fine-grained spatial temporal MLP architecture for metro OD prediction problem. Our ODMixer learns the short-term temporal relations of OD pairs by incorporating the Channel Mixer. The Multi-view Mixer efficiently captures OD pair relations from both origin and destination perspectives. With the integration of BTL, our ODMixer can perceive long-term traffic changes. Experimental results represent ODMixer’s outstanding performance on two large-scale datasets. Future directions for ODMixer involve incorporating additional city information, such as urban population distribution, regional composition, and Point of Interest (POI). Moreover, enhancing the model’s ability to learn the general pattern of traffic flow can improve its migration capability and scalability. Deploying ODMixer in actual metro systems can facilitate the management and optimization of metro operations, thereby enhancing transportation efficiency.