Conditional Monte Carlo Learning for Diffusions I: main methodology and application to backward stochastic differential equations - Archive ouverte HAL
Preprints, Working Papers, ... Year : 2020

Conditional Monte Carlo Learning for Diffusions I: main methodology and application to backward stochastic differential equations

Abstract

We present a new algorithm based on a One-layered Nested Monte Carlo (1NMC) to simulate functionals U of a Markov process X. The main originality of the proposed methodology comes from the fact that it provides a recipe to simulate U_{t≥s} conditionally on X_s. Because of the nested structure that allows a Taylor-like expansion, it is possible to use a very reduced basis for the regression. Although this methodology can be adapted for a large number of situations, we only apply it here for the simulation of Backward Stochastic Differential Equations (BSDEs). The generality and the stability of this algorithm, even in high dimension, make its strength. It is heavier than a straight Monte Carlo (MC) but it is far more accurate to simulate quantities that are almost impossible to simulate with MC. The parallel suitability of 1NMC makes it feasible in a reasonable computing time. This paper explains the main version of this algorithm and provides first results of error estimates. We also give various numerical examples with a dimension equal to 100 that are executed from few seconds to few minutes on one Graphics Processing Unit (GPU).
Fichier principal
Vignette du fichier
Cond.Learn I.pdf (825.69 Ko) Télécharger le fichier
Cond.Learn II.pdf (1.49 Mo) Télécharger le fichier
Origin Files produced by the author(s)
Origin Files produced by the author(s)
Loading...

Dates and versions

hal-02959492 , version 1 (06-10-2020)

Identifiers

  • HAL Id : hal-02959492 , version 1

Cite

Lokman Abbas-Turki, G. Pagès, Babacar Diallo. Conditional Monte Carlo Learning for Diffusions I: main methodology and application to backward stochastic differential equations. 2020. ⟨hal-02959492⟩
1350 View
317 Download

Share

More