On Hoeffding's inequality for dependent random variables
The present article investigates the relationship between pointwise probability inequalities and their uniform (parameter independent) counterparts obtained by the chaining technique. The main result of the paper is the derivation (based on Orlicz norm) of a pointwise inequality, which is shown to be a generalization of the Hoeffding's inequality. This result allows a generalization of a related bounded difference inequality of \textit{C. McDiarmid} [in: Surveys in combinatorics. Lond. Math. Soc. Lect. Note Ser. 141, 148-188 (1989; Zbl 0712.05012)]. The same result allows extension of the Hoeffding's pointwise inequality to its corresponding uniform inequality for weighted sums, involving the partitioning entropy assumption. Pointwise inequalities as well as their transitions to the uniform case are quite relevant issues for the theory of stochastic processes; for instance, the author shows how the obtained inequalities can be applied to the theory of M-estimation in the autoregression model. For the entire collection see [Zbl 1005.00016].
- Non-asymptotic tests of model performance
- Kernel Knockoffs Selection for Nonparametric Additive Models
- A nonparametric eigenvalue-regularized integrated covariance matrix estimator for asset return data
- Hoeffding's inequality for sums of dependent random variables
- On the Komlós-Révész SLLN for dependent variables
- About Doob's inequality, entropy and Tchebichef
- Stable polynomials and sums of dependent Bernoulli random variables: application to Hoeffding inequalities
- Sequential complexities and uniform martingale laws of large numbers
- Concentration of weakly dependent Banach-valued sums and applications to statistical learning methods
- On the rate of convergence of the maximum likelihood estimator in Brownian semimartingale models
- A Hoeffding-type inequality for ergodic time series
- On extensions of Hoeffding's inequality for panel data
- On Kolmogorov’s converse inequality for dependent random variables
- Data-driven model selection for same-realization predictions in autoregressive processes
- New concentration bounds and their applications in online resource allocation
- Knowledge-driven approximate model predictive controller with probabilistic performance validation
- Bernstein's inequality with a proper cover and upper bound of portfolio VaR and TVaR
- Extensions of the Hoeffding-Azuma inequalities
- Concentration inequality for evolutionary trees
- Hoeffding's inequalities: A counterexample
- Central limit theorems for the zeros of Lamé polynomials
- Locally adaptive estimation of evolutionary wavelet spectra
- An extension of the Hoeffding inequality to unbounded random variables
- Bounds for tail probabilities of martingales using skewness and kurtosis
This page was built for publication: On Hoeffding's inequality for dependent random variables
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4410077)