arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2309.09074cs.LG
收藏

面向极端交通预测的测试时补偿表示学习

Test-Time Compensated Representation Learning for Extreme Traffic Forecasting

The Hong Kong University of Science and Technology(香港科技大学)

Zhiwei Zhang, Weizhong Zhang, Yaowei Huang, Kani Chen

AI总结:

本文针对极端交通事件下多元序列预测不可靠的问题,提出测试时补偿表示学习框架CompFormer,通过时空分解数据银行与多头空间Transformer迁移历史鲁棒特征,在METR-LA和PEMS-BAY上较六个强基线最高提升28.2%。

摘要:

交通预测是一项具有挑战性的任务,原因在于交通序列之间复杂的时空相关性。在本文中,我们识别了多元交通序列预测中一个尚未被充分探索的问题:极端事件。道路拥堵和高峰时段可能导致相邻时间段内不同路口的车辆速度之间相关性较低。现有方法通常基于近期观测来预测未来序列,并在测试阶段完全丢弃训练数据,这使得它们在预测高度非线性的多元时间序列时不可靠。为了解决这一问题,我们提出了一种测试时补偿表示学习框架,该框架由一个时空分解数据银行和一个多头空间Transformer模型(CompFormer)组成。前者根据周期性特征沿时间维度显式地分离所有训练数据,而后者通过空间注意力矩阵在近期观测与数据银行中的历史序列之间建立联系。这使得CompFormer能够在消耗较少计算资源的同时迁移鲁棒特征以克服异常事件。我们的模块可以通过端到端训练灵活地与现有预测方法集成,并且我们在METR-LA和PEMS-BAY基准上证明了其有效性。大量实验结果表明,我们的方法在极端事件中尤为重要,并且相较于六个强基线能够取得显著改进,总体改进幅度最高可达28.2%。

英文摘要:

Traffic forecasting is a challenging task due to the complex spatio-temporal correlations among traffic series. In this paper, we identify an underexplored problem in multivariate traffic series prediction: extreme events. Road congestion and rush hours can result in low correlation in vehicle speeds at various intersections during adjacent time periods. Existing methods generally predict future series based on recent observations and entirely discard training data during the testing phase, rendering them unreliable for forecasting highly nonlinear multivariate time series. To tackle this issue, we propose a test-time compensated representation learning framework comprising a spatio-temporal decomposed data bank and a multi-head spatial transformer model (CompFormer). The former component explicitly separates all training data along the temporal dimension according to periodicity characteristics, while the latter component establishes a connection between recent observations and historical series in the data bank through a spatial attention matrix. This enables the CompFormer to transfer robust features to overcome anomalous events while using fewer computational resources. Our modules can be flexibly integrated with existing forecasting methods through end-to-end training, and we demonstrate their effectiveness on the METR-LA and PEMS-BAY benchmarks. Extensive experimental results show that our method is particularly important in extreme events, and can achieve significant improvements over six strong baselines, with an overall improvement of up to 28.2%.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑