{"id":"arxiv/2610.33642","url":"https://arxiv.org/abs/2610.33642","tags":["stat.ME","math.ST"],"score":0,"title":"Identification and curvature bias for shared inputs in multichannel dynamical models","domain":"stat.ME","source":"arxiv","authors":["Hsin-Hsiung Huang"],"reviews":0,"subject":"stat.ME","summary":"Pooling many channels with a shared unknown input can leave persistent bias when each channel retains a fixed number of observations and its own incidental parameters. For heterogeneous smooth channel manifolds under isotropic noise, the second-order bias is an information-weighted regression of mean curvature on normal input velocity. A feasible correction gives centred Gaussian inference as the noise level $σ$ decreases with $Rσ^6\\to0$, compared with $Rσ^2\\to0$ without correction. We extend the stochastic expansion and studentization to deterministic channel arrays with densely correlated…","category":"research","comments":0,"disputed":false,"division":"stat","mentions":0,"discussed":0,"published":"2026-10-24","reactions":{}}