Efficient Sampling of Stochastic Differential Equations with Positive Semi-Definite Models

03/30/2023
by   Anant Raj, et al.
0

This paper deals with the problem of efficient sampling from a stochastic differential equation, given the drift function and the diffusion matrix. The proposed approach leverages a recent model for probabilities <cit.> (the positive semi-definite – PSD model) from which it is possible to obtain independent and identically distributed (i.i.d.) samples at precision ε with a cost that is m^2 d log(1/ε) where m is the dimension of the model, d the dimension of the space. The proposed approach consists in: first, computing the PSD model that satisfies the Fokker-Planck equation (or its fractional variant) associated with the SDE, up to error ε, and then sampling from the resulting PSD model. Assuming some regularity of the Fokker-Planck solution (i.e. β-times differentiability plus some geometric condition on its zeros) We obtain an algorithm that: (a) in the preparatory phase obtains a PSD model with L2 distance ε from the solution of the equation, with a model of dimension m = ε^-(d+1)/(β-2s) (log(1/ε))^d+1 where 0<s≤1 is the fractional power to the Laplacian, and total computational complexity of O(m^3.5log(1/ε)) and then (b) for Fokker-Planck equation, it is able to produce i.i.d. samples with error ε in Wasserstein-1 distance, with a cost that is O(d ε^-2(d+1)/β-2log(1/ε)^2d+3) per sample. This means that, if the probability associated with the SDE is somewhat regular, i.e. β≥ 4d+2, then the algorithm requires O(ε^-0.88log(1/ε)^4.5d) in the preparatory phase, and O(ε^-1/2log(1/ε)^2d+2) for each sample. Our results suggest that as the true solution gets smoother, we can circumvent the curse of dimensionality without requiring any sort of convexity.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
12/23/2021

Asymptotic normality of least squares estimators to stochastic differential equations driven by fractional Brownian motions

We will consider the following stochastic differential equation (SDE): ...
research
04/11/2018

LAN property for stochastic differential equations driven by fractional Brownian motion of Hurst parameter H∈(1/4,1/2)

In this paper, we consider the problem of estimating the drift parameter...
research
04/26/2021

Wasserstein distance estimates for the distributions of numerical approximations to ergodic stochastic differential equations

We present a framework that allows for the non-asymptotic study of the 2...
research
03/16/2023

A Stochastic Method for Solving Time-Fractional Differential Equations

We present a stochastic method for efficiently computing the solution of...
research
04/02/2021

Projection Estimators of the Stationary Density of a Differential Equation Driven by the Fractional Brownian Motion

The paper deals with projection estimators of the density of the station...
research
11/23/2022

Model-agnostic stochastic model predictive control

We propose a model-agnostic stochastic predictive control (MASMPC) algor...
research
10/13/2020

Stochastic embeddings of dynamical phenomena through variational autoencoders

System identification in scenarios where the observed number of variable...

Please sign up or login with your details

Forgot password? Click here to reset