arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2605.08552stat.MLcs.LG
收藏

高维多组件独立成分分析中的可学习性与竞争

Learnability and Competition in High-Dimensional Multi-Component ICA

MLIP Research Group, KUIS AI Center, Koç University(MLIP研究组、KUIS人工智能中心、科克大学);Department of EEE, Koç University(电子工程系、科克大学)

Eser Ilke Genc, Samet Demir, Zafer Dogan

AI总结:

本文研究高维多组件ICA的可学习性与竞争机制,提出基于均场理论的分析,揭示初始化驱动的相结构,包括解耦和竞争阶段,并通过实验验证预测结果。

摘要:

本文研究高维多组件ICA的可学习性与竞争机制,提出基于均场理论的分析,揭示初始化驱动的相结构,包括解耦和竞争阶段,并通过实验验证预测结果。

英文摘要:

Independent Component Analysis (ICA) is a foundational tool for unsupervised representation learning, yet its high-dimensional theory remains largely limited to single-component recovery. We develop an asymptotically exact mean-field theory for multi-component online ICA, capturing the coupling induced by simultaneous learning and orthogonalization. In the high-dimensional limit, the joint empirical distribution of learned estimates and ground-truth components converges to a deterministic process, yielding a closed ODE system for the overlap matrix between learned directions and true components. This characterization reveals a genuinely multi-component, initialization-driven phase structure: a decoupled regime, where estimates align with distinct components and evolve nearly independently, and a competition regime, where overlapping initializations induce orthogonality-driven conflicts, slow reorientation, and delayed convergence. Our steady-state analysis gives explicit learnability boundaries and competition conditions linking step size, data moments, and initialization. These conditions show that larger higher-order moments and competition shrink the stable learning-rate window, increase convergence times, and predict a staircase phenomenon in which the number of recoverable components changes discretely with the learning rate. Experiments on synthetic data and hyperspectral remote sensing data validate the predicted trajectories and phase behavior.

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。

↑