Maximum entropy distribution of order statistics with given marginals (Q2405177): Difference between revisions

The authors of the paper under review consider distributions of ordered random vectors with given one-dimensional marginal distributions. Their main result here is a necessary and sufficient condition for the existence of such a unique distribution with the maximal entropy. Explicit formulae for the density function of this distribution and the value of its entropy are also provided. The core structure of the paper is as follows. In Section 3 the connection between copulas of order statistics with fixed marginals, and symmetric copulas with fixed multidiagonals are described. Next, it is shown that there exists a one-to-one map between these two sets of copulas, which has good properties with respect to the entropy. In Section 4 the authors determine the maximum entropy copula with fixed multidiagonal. In Section 5, they give the maximum entropy copula of order statistics with fixed marginals and obtain the density of the maximum entropy distribution for order statistics with fixed marginals (by composing the maximum entropy copula with these marginals). Section 6 contains the detailed proofs, while Section 7 collects the main notation of the paper to facilitate reading.