DS-Compatible Log-Linear Reliability with KL-Prox EM: Monotone Ascent, Identifiability, and Generalization

Shiva Koreddi, Sravani Sowrupilli

Keywords: truth discovery; Dawid--Skene; EM; mirror descent; proximal methods; calibration; multi-agent perception

Abstract: We study context-conditioned reliability in Dawid--Skene (DS) models and propose a DS-compatible parameterization in which a log-linear correction to confusion logits is softmax-renormalized over reported labels, yielding valid, interpretable confusion matrices conditioned on covariates. We derive a KL-proximal (mirror-descent) update for confusion matrices that warm-starts at DS and provably yields monotone ascent of a standard EM surrogate. Under diagonal-dominance at warm start and mild covariate excitation, we prove identifiability up to permutation; for the correction head we give a finite-sample generalization bound of order $O(\sqrt{d\log(dK)/n})$ via Rademacher complexity. The formulation drops into vanilla EM, preserves DS interpretability, and supports physics-inspired priors (e.g., monotonicity) without breaking guarantees. Empirical validation on multi-agent label fusion confirms monotone convergence and calibration improvements with minimal overhead.

Submission Number: 31