Multi Level Monte Carlo methods for a class of ergodic stochastic differential equations

05/04/2016
by   Lukasz Szpruch, et al.
0

We develop a framework that allows the use of the multi-level Monte Carlo (MLMC) methodology (Giles 2015) to calculate expectations with respect to the invariant measures of ergodic SDEs. In that context, we study the (over-damped) Langevin equations with strongly convex potential. We show that, when appropriate contracting couplings for the numerical integrators are available, one can obtain a time-uniform estimates of the MLMC variance in stark contrast to the majority of the results in the MLMC literature. As a consequence, one can approximate expectations with respect to the invariant measure in an unbiased way without the need of a Metropolis- Hastings step. In addition, a root mean square error of O(ϵ) is achieved with O(ϵ^-2) complexity on par with Markov Chain Monte Carlo (MCMC) methods, which however can be computationally intensive when applied to large data sets. Finally, we present a multilevel version of the recently introduced Stochastic Gradient Langevin (SGLD) method (Welling and Teh, 2011) built for large datasets applications. We show that this is the first stochastic gradient MCMC method with complexity O(ϵ^-2|ϵ|^3), which is asymptotically an order ϵ lower than the O(ϵ^-3) complexity of all stochastic gradient MCMC methods that are currently available. Numerical experiments confirm our theoretical findings.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
09/15/2016

Multilevel Monte Carlo for Scalable Bayesian Computations

Markov chain Monte Carlo (MCMC) algorithms are ubiquitous in Bayesian co...
research
03/03/2021

Full history recursive multilevel Picard approximations for ordinary differential equations with expectations

We consider ordinary differential equations (ODEs) which involve expecta...
research
02/10/2016

Stochastic Quasi-Newton Langevin Monte Carlo

Recently, Stochastic Gradient Markov Chain Monte Carlo (SG-MCMC) methods...
research
10/07/2022

Gradient-based optimisation of the conditional-value-at-risk using the multi-level Monte Carlo method

In this work, we tackle the problem of minimising the Conditional-Value-...
research
03/03/2021

Stochastic optimization for numerical evaluation of imprecise probabilities

In applications of imprecise probability, analysts must compute lower (o...
research
12/06/2021

Bounding Wasserstein distance with couplings

Markov chain Monte Carlo (MCMC) provides asymptotically consistent estim...
research
12/22/2020

Unbiased Gradient Estimation for Distributionally Robust Learning

Seeking to improve model generalization, we consider a new approach base...

Please sign up or login with your details

Forgot password? Click here to reset