Skip to main content
QUICK REVIEW

[论文解读] Coupled Layer-wise Graph Convolution for Transportation Demand Prediction

Junchen Ye, Leilei Sun|arXiv (Cornell University)|Dec 15, 2020
Traffic Prediction and Management TechniquesEngineering参考文献 34被引用 22
一句话总结

本文提出CCRNN,一种用于交通需求预测的新颖图卷积网络,采用自学习的、逐层的邻接矩阵以及耦合的逐层机制,以捕捉多层次空间依赖性和时间动态。该模型实现了最先进性能,在纽约市Citi Bike和纽约市出租车数据集上,相比Graph WaveNet分别将RMSE降低13.85%和26.85%。

ABSTRACT

Graph Convolutional Network (GCN) has been widely applied in transportation demand prediction due to its excellent ability to capture non-Euclidean spatial dependence among station-level or regional transportation demands. However, in most of the existing research, the graph convolution was implemented on a heuristically generated adjacency matrix, which could neither reflect the real spatial relationships of stations accurately, nor capture the multi-level spatial dependence of demands adaptively. To cope with the above problems, this paper provides a novel graph convolutional network for transportation demand prediction. Firstly, a novel graph convolution architecture is proposed, which has different adjacency matrices in different layers and all the adjacency matrices are self-learned during the training process. Secondly, a layer-wise coupling mechanism is provided, which associates the upper-level adjacency matrix with the lower-level one. It also reduces the scale of parameters in our model. Lastly, a unitary network is constructed to give the final prediction result by integrating the hidden spatial states with gated recurrent unit, which could capture the multi-level spatial dependence and temporal dynamics simultaneously. Experiments have been conducted on two real-world datasets, NYC Citi Bike and NYC Taxi, and the results demonstrate the superiority of our model over the state-of-the-art ones.

研究动机与目标

  • 解决现有GCN中固定、启发式邻接矩阵的局限性,这些矩阵无法捕捉真实的空间关系和交通需求中的多级依赖。
  • 通过为网络每层启用不同的邻接矩阵,实现对分层空间依赖性的建模,例如全局天气影响与局部交通事件之间的差异。
  • 通过逐层耦合机制连接高层和低层邻接矩阵,降低参数规模和计算成本。
  • 通过门控循环单元(GRU)将空间表征与时间动态相结合,实现端到端的时空预测。
  • 证明自学习邻接矩阵在真实世界需求预测中优于基于距离或相关性的初始化方法。

提出的方法

  • 提出耦合的逐层图卷积(CGC),其中每一层使用独立的可训练邻接矩阵,而非固定邻接矩阵。
  • 引入一种逐层耦合机制,通过可学习变换将高层邻接矩阵与低层邻接矩阵连接,从而减少参数数量。
  • 采用多级聚合模块,根据各层表征的相关性动态分配其重要性。
  • 将CGC的空间隐藏状态与门控循环单元(GRU)融合,以建模时间动态并生成最终的需求预测。
  • 通过反向传播端到端训练整个模型,联合优化邻接矩阵、图卷积和预测头。
  • 使用自适应嵌入(E₁⁰, E₂⁰)初始化邻接矩阵,这些嵌入在训练过程中联合学习,避免依赖于距离或相关性启发式方法。

实验结果

研究问题

  • RQ1自学习的、与层相关的邻接矩阵是否能改善交通需求预测中多层次空间依赖性的建模?
  • RQ2邻接矩阵之间的逐层耦合机制是否能在降低参数复杂度的同时保持或提升预测性能?
  • RQ3CCRNN在真实世界数据集上的性能与Graph WaveNet和DCRNN等最先进模型相比如何?
  • RQ4自适应邻接矩阵在多大程度上优于基于空间距离或皮尔逊相关系数的启发式初始化方法?
  • RQ5将空间表征与GRU结合是否能有效建模需求模式中的长期时间动态?

主要发现

  • CCRNN在纽约市Citi Bike数据集上达到最低RMSE(2.8382),在纽约市出租车数据集上达到9.5631,优于所有基线模型。
  • 与Graph WaveNet相比,CCRNN在Citi Bike和出租车数据集上分别将RMSE降低13.85%和26.85%。
  • 消融实验表明,邻接矩阵的随机初始化导致性能较差(Citi Bike上PCC = 0.0204),证实了合理初始化的必要性。
  • 基于PCC的初始化在Citi Bike上获得PCC = 0.7851,在出租车数据集上获得0.9619,优于所有基线模型,但仍逊于完整版CCRNN模型。
  • 无耦合变体在Citi Bike上PCC为0.7702,在出租车数据集上为0.9594,表明即使邻接矩阵可训练,耦合机制仍能提升性能。
  • 完整版CCRNN模型在Citi Bike上PCC为0.7934,在出租车数据集上为0.9648,证明了完整架构在捕捉复杂时空依赖关系方面的有效性。

更好的研究,从现在开始

从阅读论文到最终审阅,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。