arXiv cs.AI· Danru Xu, Lauri Parkkonen, Sara Magliacane, Aapo Hyv\"arinen·· 12 小时前AI 评分20
从多域线性混合中识别相关成分：MuDo-CoM 算法
Identifying Dependent Components from Multi-domain Linear Mixtures
AI 导读
研究提出多域线性混合模型，允许潜在成分之间存在依赖关系，各域仅通过成分的域特定缩放不同，而潜在成分分布跨域一致。在域变异性充分时，仅凭二阶统计量即可识别潜在变量与混合函数，据此提出的 Multi-Domain Covariance Matching（MuDo-CoM）算法推广了此前的联合对角化方法。MuDo-CoM 已在模拟数据和真实 fMRI 数据集上得到验证。
正文 · 原文
Abstract:We study a linear mixing model with dependent latent components, assuming multiple data domains. Most existing models assume that the components are independent or at least uncorrelated, in line with independent component analysis (ICA). Some recent work allows for dependent components, but then makes specific assumptions such as parametric forms of dependencies, multi-view settings, or interventions, or does not recover the individual components. In contrast, we consider a multi-domain setting in which domains differ through domain-specific scalings of the components, while the distribution of the underlying latent components is the same across domains. This approach can model data collected, for example, from different sensors measuring the same process, different laboratories conducting the same experiment, different experimental conditions, or different subjects that might differ in biological or physiological factors. We show that, under sufficient domain variability, latent variables and mixing functions can be identified from second-order statistics alone. We propose the Multi-Domain Covariance Matching (MuDo-CoM) algorithm that generalizes previous methods of joint diagonalization. MuDo-CoM is validated on simulated data and a real-world fMRI dataset.
| Comments: | 29 pages, 3 figures |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.28330 [cs.LG] |
| (or arXiv:2610.28330v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.28330 arXiv-issued DOI via DataCite (pending registration) |
|
| Journal reference: | Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:141600-141628, 2026 |
Submission history
From: Danru Xu [view email]
[v1]
Tue, 20 Oct 2026 15:01:03 UTC (1,371 KB)
来源：arXiv cs.AI · arxiv.org