Approximation of heavy-tailed distributions via stable-driven SDEs
From MaRDI portal
Publication:2040106
Abstract: Constructions of numerous approximate sampling algorithms are based on the well-known fact that certain Gibbs measures are stationary distributions of ergodic stochastic differential equations (SDEs) driven by the Brownian motion. However, for some heavy-tailed distributions it can be shown that the associated SDE is not exponentially ergodic and that related sampling algorithms may perform poorly. A natural idea that has recently been explored in the machine learning literature in this context is to make use of stochastic processes with heavy tails instead of the Brownian motion. In this paper we provide a rigorous theoretical framework for studying the problem of approximating heavy-tailed distributions via ergodic SDEs driven by symmetric (rotationally invariant) -stable processes.
Heavy-tailed distributions have been found to be very useful in many applications. Yet related sampling algorithms may perform poorly. In this paper, the authors provide a rigorous theoretical framework for studying the problem of approximating heavy-tailed distributions via ergodic SDEs driven by rotationally invariant \(\alpha\)-stable processes.
Recommendations
- scientific article; zbMATH DE number 1995730
- Distribution tails for solutions of SDE driven by an asymmetric stable Lévy process
- Heavy-tailed Distributions in Some Stochastic Dynamical Models
- scientific article; zbMATH DE number 1221798
- Remarks on the rate of strong convergence of Euler-Maruyama approximation for SDEs driven by rotation invariant stable processes
Cites work
- A note on existence of global solutions and invariant measures for jump SDEs with locally one-sided Lipschitz drift
- Analysis and geometry of Markov diffusion operators
- Coupling property and gradient estimates of Lévy processes via the symbol
- Dirichlet heat kernel estimates for fractional Laplacian with gradient perturbation
- Eigenvalues, Inequalities, and Ergodic Theory
- Ergodicity of stochastic differential equations with jumps and singular coefficients
- Exponential convergence of Langevin distributions and their discrete approximations
- Functional inequalities for stable-like Dirichlet forms
- Heat kernel estimates for critical fractional diffusion operators
- Heat kernels for time-dependent non-symmetric stable-like operators
- scientific article; zbMATH DE number 3329342 (Why is no real title available?)
- Lévy matters III. Lévy-type processes: construction, approximation and sample path properties
- Nonasymptotic bounds for sampling algorithms without log-concavity
- Nonasymptotic convergence analysis for the unadjusted Langevin algorithm
- Nonlocal diffusion and applications
- On explicit approximations for Lévy driven SDEs with super-linear diffusion coefficients
- On multidimensional Ornstein-Uhlenbeck processes driven by a general Lévy process
- On stochastic equations with respect to semimartingales I.†
- On tamed Euler approximations of SDEs driven by Lévy noise with applications to delay equations
- On the rate of convergence of strong Euler approximation for SDEs driven by Levy processes
- On weak uniqueness and distributional properties of a solution to an SDE with -stable noise
- Pathwise uniqueness for singular SDEs driven by stable processes
- Potential analysis of stable processes and its extensions. With some of the papers based on the presentations at the workshop on stochastic and harmonic analysis of processes with jumps, Angers, France, May 2--9, 2006.
- Quantitative contraction rates for Markov chains on general state spaces
- Quantitative Harris-type theorems for diffusions and McKean-Vlasov processes
- Reflection couplings and contraction rates for diffusions
- Riesz potential operators and inverses via fractional centred derivatives
- Stability of Markovian processes II: continuous-time processes and sampled chains
- Stability of Markovian processes III: Foster–Lyapunov criteria for continuous-time processes
- Stochastic flows for Lévy processes with Hölder drifts
- Stochastic stability of differential equations. With contributions by G. N. Milstein and M. B. Nevelson
- Strong convergence of the Euler-Maruyama approximation for a class of Lévy-driven SDEs
- Ten equivalent definitions of the fractional Laplace operator
- Theoretical Guarantees for Approximate Sampling from Smooth and Log-Concave Densities
Cited in
(9)- Exponential ergodicity for SDEs and McKean-Vlasov processes with Lévy noise
- Strict Kantorovich contractions for Markov chains and Euler schemes with general noise
- Simple nonlinear models with rigorous extreme events and heavy tails
- Ergodicity of supercritical SDEs driven by \(\alpha \)-stable processes and heavy-tailed sampling
- Lévy Langevin Monte Carlo
- Explicit results for ergodic properties of SDEs driven by cylindrical symmetric stable noise
- L^2-exponential ergodicity of stochastic Hamiltonian systems with -stable Lévy noises
- Convergence rates for inhomogeneous Markov chains from stochastic approximation algorithms
- Stationary distributions of McKean-Vlasov SDEs with jumps: existence, multiplicity and uniqueness
This page was built for publication: Approximation of heavy-tailed distributions via stable-driven SDEs
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2040106)
