[Submitted on 18 Oct 2026 (v1), last revised 23 Oct 2026 (this version, v2)]
Abstract:Strict replicability requires two independent training samples, processed with the same random seed, to produce exactly the same hypothesis. We study realizable Boolean learning under unknown and known instance marginals. For unknown marginals, we obtain the worst-case sample complexity $Theta(((log M)^2/rho^2+log(1/delta))/varepsilon)$ for classes of cardinality at most $M$, throughout a fixed small-parameter range. This resolves the distribution-free finite-class sample-complexity question up to universal constants and rules out a general $rho^{-1}$ dependence. The upper bound uses a truncated-power distribution with a hard empirical-error budget. For known marginals, a control-variate learner uses $O(((1+log M)^{3/2}/rho+log(1/delta))/varepsilon)$ samples, and explicit uniform-marginal instances establish the matching class-size exponent. On a product of $d$ finite threshold chains of length $k$, we determine the exact rate $Theta((dklog(1+sqrt d/(krho))+log(1/delta))/varepsilon)$. We also characterize deterministic learning, the very-small-$rho$ regime for arbitrary finite classes, and every cardinality-replicability regime for finite classes of VC dimension one. A VC-covering-number bound yields the fixed-accuracy exponent $3/2$ independently of cardinality. Finally, we separate statistical complexity from implementation costs: we prove a sharp fixed-block trade-off, short-seed existence bounds, and a shared rejection-filter reduction from optimal learning of the full Boolean class to efficient public sampling of ordinary Bernoulli product distributions. Detailed proofs include the geometric lower bounds, the discrete posterior-kernel comparison, and the finite-bit constructions.
Submission history
From: Guangjian Zhang [view email]
[v1] Sun, 18 Oct 2026 05:17:14 UTC (27 KB)
[v2] Fri, 23 Oct 2026 01:18:20 UTC (67 KB)
