Optimal tail comparison based on comparison of moments
Ce problème est traité comme une recherche d'extremum sous contrainte: \( F\) étant un ensemble de fonctions réelles sur un espace mesurable \((K, \Sigma)\), on cherche à maximiser \(\mathbb{E}\psi(\xi)\) sous la contrainte \(\mathbb{E} \varphi (\xi)\leq\mathbb{E}\varphi (\eta)\) pour tout \(\varphi \in F\), \(\psi\) étant une fonction sur \(K\), \(\xi\) et \(\eta\) des v.a. sur un espace mesurable \((K, \Sigma)\). Utilisant une propriété de dualité remontant à Chebyshev, la question revient à l'évaluation de: \(I(\nu,\psi, \alpha)= \inf \{\int\varphi d\nu: \varphi\geq \psi,\;\varphi\in F_\alpha\}\), où \(\nu\) est la mesure de la v.a. \(\eta\) tandis que \(F_\alpha\) sont des fonctions mesurables \(\alpha\)-convexes (càd non décroissantes et telles que: \((\varphi-\varphi (-\infty))^{1/ \alpha}\) soit convexe sur \(\mathbb{R})\). En particulier, les bornes supérieures précises de \(I\) sont déterminées pour des queues de distribution de la mesure \(\nu\), ainsi que des propriétés asymptotiques. Application aux vecteurs gaussiens, à des intégrales stochastiques et semi-martingales. For the entire collection see [Zbl 0883.00024].
- Necessary conditions of extremality of extreme distributions in generalized Chebyshev inequalities and their application to reliability theory. I
- Sufficient conditions of extremality of extreme distributions \(F_{5} - F_{7}\) in generalized Chebyshev inequalities. III
- Extremal values of a class of functionals with moment constraints
- Necessary and sufficient conditions of existence of extreme distributions \(F_{5}-F_{7}\) in the generalized Chebyshev inequalities problem. II.
- scientific article; zbMATH DE number 3986024
- Bounds for tail probabilities of the sample variance
- Comparison between two indicators for the variation regularity of tails of distributions
- Comparison inequalities for suprema of bounded empirical processes
- Concentration inequalities for separately convex functions
- Positive-part moments via characteristic functions, and more general expressions
- On the dependence of the Berry-Esseen bound on dimension
- On Hoeffding's inequalities.
- Probability inequalities for bounded random vectors
- An inequality for the sum of independent bounded random variables
- An exponential inequality for suprema of empirical processes with heavy tails on the left
- On the Bennett-Hoeffding inequality
- Exact inequalities for sums of asymmetric random variables, with applications
- Exact lower bounds on the exponential moments of truncated random variables
- Exact Bounds on the Truncated-Tilted Mean, with Applications
- Exact converses to a reverse AM-GM inequality, with applications to sums of independent random variables and (super)martingales
- Rademacher-Gaussian tail comparison for complex coefficients and related problems
- Toward the best constant factor for the Rademacher-Gaussian tail comparison
- Positive-part moments via the Fourier-Laplace transform
- On Gaussian and Bernoulli covariance representations
- Addendum to ``An extension of an inequality of Hoeffding to unbounded random variables: The non-i.i.d. case
- A sharp Gaussian tail bound for sums of uniforms
- Exact Rosenthal-type bounds
- On domination of tail probabilities of (super)martingales: explicit bounds
- An extension of the Hoeffding inequality to unbounded random variables
- Bounds for tail probabilities of martingales using skewness and kurtosis
- On measure concentration for separately Lipschitz functions in product spaces
This page was built for publication: Optimal tail comparison based on comparison of moments
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4387152)