Contextual features are important data sources for building spatiotemporal crowd flow prediction (STCFP) models. However, the difficulty of applying context lies in the unknown generalizability of both contextual features (e.g., weather, holiday, and points-of-interests) and context modeling techniques across different scenarios. In this paper, we develop an experimental platform composed of large-scale spatiotemporal crowd flow data, contextual data, and state-of-the-art spatiotemporal prediction models to conduct a comprehensive experimental study to quantitatively investigate the generalizability of different contextual features and modeling techniques in three urban crowd flow prediction scenarios (bike flow, metro passenger flow, and electric vehicle charging demand). In particular, we develop a general taxonomy of context modeling techniques based on extensive investigations in prevailing research. With three real-world datasets including millions of records and rich context data, we have trained and tested hundreds of different models. Our results reveal several important observations: (1) Using more contextual features may not always result in better prediction with existing context modeling techniques; in particular, the contextual feature combination of holiday and temporal position can provide more generalizable beneficial information than other contextual feature combinations. (2) In context modeling techniques, using a gated unit to incorporate raw contextual features into the state-of-the-art prediction model has good generalizability. Besides, we also offer several suggestions about incorporating contextual factors for practitioners who want to build STCFP applications. From our findings, we call for future research efforts devoted to developing new context processing and modeling solutions to fully exploit the potential of contextual features for STCFP.
翻译:在本文中,我们开发了一个实验平台,由大型超时人群流动数据、相关数据和最先进的超时人群流动预测模型组成,以进行综合实验研究,从数量上调查不同背景特征和模型技术在三种城市人群流动预测假设情景(即电动流动、地铁客流和电动车辆充电需求)中的可概括性,但是,背景应用的困难在于背景特征(如天气、假日和利益点)和背景建模技术在不同情景中的可知性。在本文中,我们开发了一个实验平台,由大型超时人群流动数据、背景数据和最先进的超时空预测模型组成,以进行一项全面的实验性研究,以从数量上调查不同背景特征和模型技术在三种城市人群流动预测情景预测情景预测情景预测情景中的可概括性(如电动流动、地铁客流和电动车辆充电动车辆需求)中的可概括性。 特别是,我们根据当前研究的广泛调查,开发了背景模型数据集,包括数百万个记录和丰富的背景数据,我们培训和测试了数百个不同的模型。 我们的ST模型显示了若干重要观察结果:(1) 使用更多的背景特征,用更多的背景特征来更好地预测现有背景技术;从背景背景背景模型,从背景背景模型到背景背景背景背景背景变量组合中,我们可以充分利用一些背景模型,我们提供一些背景模型,从而提供更有利于性数据,我们提供一些背景信息。