Abstract
We exhibit solutions of Monge-Kantorovich mass transportation problems with constraints on the support of the feasible transportation plans and additional capacity restrictions. The Hoeffding-Fréchet inequalities are extended for bivariate distribution functions having fixed marginal distributions and satisfying additional constraints. Sharp bounds for different probabilistic functionals (e.g. L p-distances, covariances, etc.) are given when the family of joint distribution functions has prescribed marginal distributions, satisfies restrictions on the support, and is bounded from above, or below, by other distributions.
Original language | English |
---|---|
Pages (from-to) | 433-445 |
Number of pages | 13 |
Journal | Journal of Applied Probability |
Volume | 36 |
Issue number | 2 |
DOIs | |
State | Published - Jun 1999 |
Keywords
- Greedy algorithms
- Linear programming
- Marginal distributions
- Measures of dependence
- Monge-Kantorovich problem
- Multivariate distributions