首页 > AI前沿 > Selective State-Space Neural Tucker Factorization for Temporal-aware Traffic Imputation

Selective State-Space Neural Tucker Factorization for Temporal-aware Traffic Imputation

arXiv机器学习 2026-11-11 22:00 16 阅读 查看原文
用户要求将一段纯文本新闻正文排版成语义化的HTML片段。 规则: 1. 只输出HTML片段本身,不要Markdown围栏、不要解释。 2. 每个自然段落用`

`包裹,长文用`

`/`

`划分小节。 3. 列表内容用`
  • `,引用用`
    `,代码用`
    `,重要句子可用``。
    4. 保留原文所有信息,不得增删改内容。
    5. 不要输出``/``/``等文档级标签。
    
    正文内容:
    Spatiotemporal traffic data imputation is a key task of intelligent transportation systems (ITS), and urban traffic data inherently exhibits strong long-term temporal dependencies. Therefore, accurate modeling of these long-term temporal dynamics is the core prerequisite for high fidelity traffic data imputation. However, existing traffic data imputation methods still suffer from inherent limitations in long-sequence modeling, such as gradient vanishing, low efficiency, and insufficient capability to capture long-range temporal correlations. To address these issues, this paper proposes a Selective State Space Neural Tucker Factorization model (S3NTF) for robust traffic data imputation. This model combines the Selective State Space mechanism with Neural Tucker Factorization to capture complex spatial correlations and long-term temporal dynamics. A state space-based time encoder is proposed to effectively extracts long sequence dependencies with linear complexity, while the latent spatiotemporal feature interaction learning module models high-order cross modal interactions based on Tucker decomposition. Extensive experiments on six real traffic datasets have shown that our model outperforms existing models in terms of accuracy.
    
    分析正文:
    这是一段学术摘要/新闻,没有明显的列表、引用或代码。可以将其划分为几个小节,或者就作为几个段落。
    为了符合“长文用`

    `/`

    `划分小节”,我可以添加一些小节标题,但规则4要求“不得增删改内容”。如果添加标题,算不算“增”内容? “保留原文所有信息,不得增删改内容”通常指不改变正文文本本身,但添加HTML结构标签(如h2/h3)是排版的一部分。不过,如果原文没有小标题,强行加小标题可能会被视为“增”内容。 更安全的做法是:把原文的几个句子/逻辑块分成几个`

    `,或者用一个`

    `作为总标题(但原文没有标题,加标题算增吗?最好不加,或者用h3表示小节,但不加文字?不行,h2/h3必须有文字)。 仔细看规则2:“每个自然段落用`

    `包裹,长文用`

    `/`

    `划分小节”。如果原文没有小节标题,就只用`

    `即可。或者把原文的几句话分成几个自然段,每个用`

    `包裹。 原文是一整段。我可以把它分成几个自然段,每段用`

    `包裹。 第一段:Spatiotemporal traffic data imputation is a key task of intelligent transportation systems (ITS), and urban traffic data inherently exhibits strong long-term temporal dependencies. Therefore, accurate modeling of these long-term temporal dynamics is the core prerequisite for high fidelity traffic data imputation. 第二段:However, existing traffic data imputation methods still suffer from inherent limitations in long-sequence modeling, such as gradient vanishing, low efficiency, and insufficient capability to capture long-range temporal correlations. 第三段:To address these issues, this paper proposes a Selective State Space Neural Tucker Factorization model (S3NTF) for robust traffic data imputation. This model combines the Selective State Space mechanism with Neural Tucker Factorization to capture complex spatial correlations and long-term temporal dynamics. 第四段:A state space-based time encoder is proposed to effectively extracts long sequence dependencies with linear complexity, while the latent spatiotemporal feature interaction learning module models high-order cross modal interactions based on Tucker decomposition. 第五段:Extensive experiments on six real traffic datasets have shown that our model outperforms existing models in terms of accuracy. 对于“重要句子可用``”,可以将核心模型名称或结论加粗。 例如:

    Spatiotemporal traffic data imputation is a key task of intelligent transportation systems (ITS), and urban traffic data inherently exhibits strong long-term temporal dependencies. Therefore, accurate modeling of these long-term temporal dynamics is the core prerequisite for high fidelity traffic data imputation.

    However, existing traffic data imputation methods still suffer from inherent limitations in long-sequence modeling, such as gradient vanishing, low efficiency, and insufficient capability to capture long-range temporal correlations.

    To address these issues, this paper proposes a Selective State Space Neural Tucker Factorization model (S3NTF) for robust traffic data imputation. This model combines the Selective State Space mechanism with Neural Tucker Factorization to capture complex spatial correlations and long-term temporal dynamics.

    A state space-based time encoder is proposed to effectively extracts long sequence dependencies with linear complexity, while the latent spatiotemporal feature interaction learning module models high-order cross modal interactions based on Tucker decomposition.

    Extensive experiments on six real traffic datasets have shown that our model outperforms existing models in terms of accuracy.

    检查是否增删改内容: 原文: Spatiotemporal traffic data imputation is a key task of intelligent transportation systems (ITS), and urban traffic data inherently exhibits strong long-term temporal dependencies. Therefore, accurate modeling of these long-term temporal dynamics is the core prerequisite for high fidelity traffic data imputation. However, existing traffic data imputation methods still suffer from inherent limitations in long-sequence modeling, such as gradient vanishing, low efficiency, and insufficient capability to capture long-range temporal correlations. To address these issues, this paper proposes a Selective State Space Neural Tucker Factorization model (S3NTF) for robust traffic data imputation. This model combines the Selective State Space mechanism with Neural Tucker Factorization to capture complex spatial correlations and long-term temporal dynamics. A state space-based time encoder is proposed to effectively extracts long sequence dependencies with linear complexity, while the latent spatiotemporal feature interaction learning module models high-order cross modal interactions based on Tucker decomposition. Extensive experiments on six real traffic datasets have shown that our model outperforms existing models in terms of accuracy. 分段后: 1. Spatiotemporal traffic data imputation is a key task of intelligent transportation systems (ITS), and urban traffic data inherently exhibits strong long-term temporal dependencies. Therefore, accurate modeling of these long-term temporal dynamics is the core prerequisite for high fidelity traffic data imputation. 2. However, existing traffic data imputation methods still suffer from inherent limitations in long-sequence modeling, such as gradient vanishing, low efficiency, and insufficient capability to capture long-range temporal correlations. 3. To address these issues, this paper proposes a Selective State Space Neural Tucker Factorization model (S3NTF) for robust traffic data imputation. This model combines the Selective State Space mechanism with Neural Tucker Factorization to capture complex spatial correlations and long-term temporal dynamics. 4. A state space-based time encoder is proposed to effectively extracts long sequence dependencies with linear complexity, while the latent spatiotemporal feature interaction learning module models high-order cross modal interactions based on Tucker decomposition. 5. Extensive experiments on six real traffic datasets have shown that our model outperforms existing models in terms of accuracy. 完全保留了原文的字词,包括拼写错误("extracts" 原文是 "extracts",保留原样)。 没有添加 h2/h3,因为原文没有小节标题,强行加标题会违反“不得增内容”的规则。用多个 `

    ` 分段即可。 加粗了重要句子,符合规则3。 没有列表、引用、代码,所以不需要相关标签。 没有文档级标签。 没有 Markdown 围栏。 只输出 HTML 片段。 最终输出:

    Spatiotemporal traffic data imputation is a key task of intelligent transportation systems (ITS), and urban traffic data inherently exhibits strong long-term temporal dependencies. Therefore, accurate modeling of these long-term temporal dynamics is the core prerequisite for high fidelity traffic data imputation.

    However,