arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.25530stat.MLcs.NAmath.NA
收藏

自适应子空间建模与功能Tucker分解

Adaptive Subspace Modeling With Functional Tucker Decomposition

KU Leuven(荷语鲁汶大学)

Noah Steidle, Joppe De Jonghe, Mariya Ishteva

AI总结:

本文提出功能Tucker分解,通过RKHS建模连续模式,实现自适应子空间建模,应用于超光谱成像和多变量时间序列分类。

摘要:

张量为多维数据提供了结构化表示,但离散化可能在数据源自连续过程时掩盖重要信息。本文通过引入功能Tucker分解(FTD),将模式间的连续性约束直接嵌入分解中。FTD利用再生核希尔伯特空间(RKHS)建模连续模式,无需先验基,同时保持Tucker模型的多线性子空间结构。通过RKHS驱动的表示,模型产生自适应且表达性强的因子描述,实现子空间的定向建模。该方法的价值在领域变异张量分类中得到验证。特别是,通过超光谱成像和多变量时间序列分析的分类任务,展示了结合结构分解与功能适应性的优势。

英文摘要:

Tensors provide a structured representation for multidimensional data, yet discretization discards the underlying continuous structure when the data originate from continuous processes. We address this limitation by introducing a functional Tucker decomposition (FTD) that embeds a mode-wise continuity constraint directly into the factorization. The FTD models the continuous mode as a function in a reproducing kernel Hilbert space (RKHS), avoiding a prespecified basis while preserving the multilinear subspace structure of the Tucker model. We derive a reconstruction error bound for the continuous mode that quantifies the approximation quality when a subspace estimated on one domain is reused on another. This bound provides theoretical justification for subspace transfer, whose practical value we demonstrate on cross-domain classification tasks in hyperspectral imaging and multivariate time-series analysis.

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。

↑