Sub-Weibull distributions: generalizing sub-Gaussian and sub-Exponential properties to heavier-tailed distributions

Abstract

We propose the notion of sub-Weibull distributions, which are characterised by tails lighter than (or equally light as) the right tail of a Weibull distribution. This novel class generalises the sub-Gaussian and sub-Exponential families to potentially heavier-tailed distributions. Sub-Weibull distributions are parameterized by a positive tail index $\theta$ and reduce to sub-Gaussian distributions for $\theta =1/2$ and to sub-Exponential distributions for $\theta =1$. A characterisation of the sub-Weibull property based on moments and on the moment generating function is provided and properties of the class are studied. An estimation procedure for the tail parameter is proposed and is applied to an example stemming from Bayesian deep learning.

arXiv comment10 pages, 3 figures

References in corpus (5)

• Deep Neural Networks as Gaussian Processes
• On the Impact of the Activation Function on Deep Neural Networks Training
• Concentration inequalities in the infinite urn scheme for occupancy counts and the missing mass, with applications
• Understanding Priors in Bayesian Neural Networks at the Unit Level
• Bootstrapping Upper Confidence Bound

Cited by in corpus (7)

• Sharper Sub-Weibull Concentrations
• Feedback-Based Optimization with Sub-Weibull Gradient Errors and Intermittent Updates
• Almost Surely $\sqrt{T}$ Regret for Adaptive LQR
• A Stochastic Operator Framework for Optimization and Learning with Sub-Weibull Errors
• Subgeometrically ergodic autoregressions with autoregressive conditional heteroskedasticity
• On Unifying Randomized Methods For Inverse Problems
• Online Stochastic Gradient Methods Under Sub-Weibull Noise and the Polyak-Łojasiewicz Condition