A Simple State Space Model Excels at Multivariate Time Series Classification
A Simple State Space Model Excels at Multivariate Time Series Classification
简单的状态空间模型在多变量时间序列分类中表现优异
Abstract: Structured state space models (SSMs) have recently emerged as a promising foundation for sequence modeling, with Mamba-based architectures demonstrating strong performance through input-dependent state transitions, albeit at considerable complexity. However, their application to time-series classification (TSC) has been largely limited to Mamba-style architectures, leaving the broader SSM design space underexplored.
摘要： 结构化状态空间模型（SSMs）最近已成为序列建模的一种有前景的基础架构。其中，基于 Mamba 的架构通过输入依赖的状态转换展现出了强大的性能，但同时也伴随着相当高的复杂性。然而，它们在时间序列分类（TSC）中的应用在很大程度上局限于 Mamba 风格的架构，这使得更广泛的 SSM 设计空间尚未得到充分探索。
We present the first systematic study spanning diagonal SSMs (S4D) and input-dependent SSMs (Mamba family) on large-scale TSC benchmarks, asking whether such complexity is necessary for top performance. Our results reveal a surprising finding: S4D consistently outperforms Mamba-based variants in both accuracy and efficiency, challenging the assumption that increased complexity translates to meaningful gains in TSC.
我们对大规模 TSC 基准测试中的对角 SSM（S4D）和输入依赖型 SSM（Mamba 系列）进行了首次系统性研究，旨在探讨这种复杂性对于实现顶级性能是否必要。我们的研究结果揭示了一个令人惊讶的发现：S4D 在准确性和效率上均持续优于基于 Mamba 的变体，这挑战了“增加复杂性即意味着在 TSC 中获得显著收益”这一假设。
Building on this, we introduce MS4, lightweight modifications to S4D via a linear input projection and channel-mixing mechanism, and MS4N, a normalized variant that stabilizes state dynamics with negligible overhead. Evaluated on 59 datasets across MONSTER (up to 60 million samples, 50K timesteps, 82 classes) and the UEA benchmark, against 15 baselines, MS4 and MS4N consistently outperform Mamba-based models while remaining more efficient, and MS4N matches or surpasses competing deep learning models that are roughly 2x and 10x larger in parameters.
在此基础上，我们引入了 MS4——通过线性输入投影和通道混合机制对 S4D 进行的轻量级修改；以及 MS4N——一种归一化变体，它能以极小的开销稳定状态动态。在涵盖 MONSTER（多达 6000 万个样本、5 万个时间步长、82 个类别）和 UEA 基准测试的 59 个数据集上，对比 15 个基准模型后发现，MS4 和 MS4N 在保持更高效率的同时，持续优于基于 Mamba 的模型；此外，MS4N 的性能与参数量约为其 2 倍至 10 倍的竞争深度学习模型相当或更胜一筹。
These results position lightweight structured SSMs as a compelling alternative to scaling complexity for TSC.
这些结果表明，轻量级结构化 SSM 是替代通过增加复杂性来提升 TSC 性能的一种极具吸引力的方案。