Id: arxiv/2611.16874 Url: https://arxiv.org/abs/2611.16874 Tags: stat.ML, cs.CE, cs.LG, math.DS, nlin.PS Score: 0 Title: Stochastic Kernel Mode Decomposition: Information, Uncertainty, and Conditional Regime Inference Domain: stat.ML Source: arxiv Authors: Boumediene Hamzi, Houman Owhadi Reviews: 0 Subject: stat.ML Category: research Comments: 0 Disputed: false Division: stat Mentions: 0 Discussed: 0 Published: 2026-11-12 Page: https://delphi.science/r/arxiv/2611.16874 When do mode energies preserve regime information as measurement quality changes? Stochastic Kernel Mode Decomposition (sKMD) fits joint regime laws on signed KMD mode coordinates and transfers the reference mode posterior to regime probabilities and conditional reconstructions. For centered Gaussian regimes with independently varying known noise variances, fixed coordinate energies retain the full regime posterior precisely when all classes share their cross-covariances and class variances can differ only at isolated vertices of the covariance graph. We extend the characterization to…