Reservoir Subspace Injection for Online ICA under Top-n Whitening
2026-03-02 • Machine Learning
Machine LearningArtificial Intelligence
AI summaryⓘ
The authors study how adding extra features (called 'reservoir expansion') can help a method called online independent component analysis (ICA) when dealing with complex, nonlinear data mixes. They identify a problem they name 'reservoir subspace injection' (RSI), where adding features may harm performance by interfering with important directions that should be kept. By creating measures to detect this problem, the authors design a controller that avoids losing these key directions and thus recovers good performance. Their method improves results compared to the standard online ICA, especially on challenging nonlinear data.
Independent Component Analysis (ICA)Reservoir ExpansionNonlinear MixingWhiteningSubspace InjectionPassthrough DirectionsSI-SDREigenspaceOnline Learning
Authors
Wenjun Xiao, Yuda Bi, Vince D Calhoun
Abstract
Reservoir expansion can improve online independent component analysis (ICA) under nonlinear mixing, yet top-$n$ whitening may discard injected features. We formalize this bottleneck as \emph{reservoir subspace injection} (RSI): injected features help only if they enter the retained eigenspace without displacing passthrough directions. RSI diagnostics (IER, SSO, $ρ_x$) identify a failure mode in our top-$n$ setting: stronger injection increases IER but crowds out passthrough energy ($ρ_x: 1.00\!\rightarrow\!0.77$), degrading SI-SDR by up to $2.2$\,dB. A guarded RSI controller preserves passthrough retention and recovers mean performance to within $0.1$\,dB of baseline $1/N$ scaling. With passthrough preserved, RE-OICA improves over vanilla online ICA by $+1.7$\,dB under nonlinear mixing and achieves positive SI-SDR$_{\mathrm{sc}}$ on the tested super-Gaussian benchmark ($+0.6$\,dB).