Scope and Construction of Balanced Mixture Covers

Investigate which distribution classes admit balanced covers of the type used to combine randomized fixed-schedule component laws, how such covers can be found from estimated conditional distributions, and how the results change when the round budget is specified in expectation rather than as a hard cap.

Background

The paper develops two finite constructions in which hidden randomization over fixed schedules combines component output laws that cover different selector-conditioned portions of a target distribution. A density-coverage lemma formalizes when such mixtures approximate the target, but the constructions rely on known finite dependence structures and exact conditional marginals.

The discussion identifies unresolved extensions concerning the generality of distributions admitting balanced covers, algorithmic discovery of suitable covers when conditional distributions are only estimated, and the effect of replacing a pathwise hard round cap with an expected round-budget constraint. These issues are not resolved by the paper's explicit finite constructions or perturbation analysis.