arXivDaily arXiv每日学术速递 周一至周五更新
北美退伍军人节(Veterans Day),12 日无更新
arXiv 2409.05933cs.LG
收藏

基于 eKAN 网络的自监督状态空间模型用于实时交通事故预测

Self-Supervised State Space Model for Real-Time Traffic Accident Prediction Using eKAN Networks

Xin Tan, Meng Zhao

AI总结:

针对交通事故预测中跨城市泛化差与实时计算成本高的问题,提出 SSL-eKamba 自监督框架,结合时空差异辅助任务与高效 eKamba/Selective SSM,在两个真实数据集上持续超越基线。

摘要:

跨不同时间和区域准确预测交通事故对公共安全至关重要。然而,现有方法面临两个关键挑战:1)泛化能力:当前模型严重依赖人工构建的多视图结构,如 POI 分布和道路网络密度,这些结构劳动密集且难以跨城市扩展。2)实时性能:尽管一些方法通过复杂架构提高了准确性,但它们通常会产生高昂的计算成本,从而限制了其实时适用性。为应对这些挑战,我们提出了 SSL-eKamba,这是一种用于交通事故预测的高效自监督框架。为增强泛化能力,我们设计了两个自监督辅助任务,通过时空差异感知自适应地改进交通模式表示。针对实时性能,我们引入了 eKamba,这是一种重新设计 Kolmogorov-Arnold Network(KAN)架构的高效模型。这包括使用可学习的一元函数进行输入激活,以及应用选择性机制(Selective SSM)来捕获多变量相关性,从而提高计算效率。在两个真实世界数据集上进行的大量实验表明,SSL-eKamba 始终优于最先进的基线方法。该框架还可能为其他时空任务提供新的见解。我们的源代码可在 http://github.com/KevinT618/SSL-eKamba 公开获取。

英文摘要:

Accurate prediction of traffic accidents across different times and regions is vital for public safety. However, existing methods face two key challenges: 1) Generalization: Current models rely heavily on manually constructed multi-view structures, like POI distributions and road network densities, which are labor-intensive and difficult to scale across cities. 2) Real-Time Performance: While some methods improve accuracy with complex architectures, they often incur high computational costs, limiting their real-time applicability. To address these challenges, we propose SSL-eKamba, an efficient self-supervised framework for traffic accident prediction. To enhance generalization, we design two self-supervised auxiliary tasks that adaptively improve traffic pattern representation through spatiotemporal discrepancy awareness. For real-time performance, we introduce eKamba, an efficient model that redesigns the Kolmogorov-Arnold Network (KAN) architecture. This involves using learnable univariate functions for input activation and applying a selective mechanism (Selective SSM) to capture multi-variate correlations, thereby improving computational efficiency. Extensive experiments on two real-world datasets demonstrate that SSL-eKamba consistently outperforms state-of-the-art baselines. This framework may also offer new insights for other spatiotemporal tasks. Our source code is publicly available at http://github.com/KevinT618/SSL-eKamba.

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。

↑