Abstract
We show that a result from combinatorial optimization theory can be applied to describe the set of all attainable mean distribution respectively classification vectors in some classes of multiple distribution and classification problems. As a special case we obtain a probabilistic version of the marriage theorem of Halmos and Vaughan. We indicate some relations to the study of robust statistical models.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have