Computation of conditional expectations with guarantees

DOI10.1007/S10915-023-02130-8arXiv2112.01804OpenAlexW4321254568MaRDI QIDQ6159022FDOQ6159022

Authors: Patrick Cheridito, Balint Gersey 

Publication date: 20 June 2023

Published in: Journal of Scientific Computing (Search for Journal in Brave)

Abstract: Theoretically, the conditional expectation of a square-integrable random variable $Y$ given a $d$-dimensional random vector $X$ can be obtained by minimizing the mean squared distance between $Y$ and $f(X)$ over all Borel measurable functions $fcolonmathbb{R}^{d}omathbbR$. However, in many applications this minimization problem cannot be solved exactly, and instead, a numerical method which computes an approximate minimum over a suitable subfamily of Borel functions has to be used. The quality of the result depends on the adequacy of the subfamily and the performance of the numerical method. In this paper, we derive an expected value representation of the minimal mean squared distance which in many applications can efficiently be approximated with a standard Monte Carlo average. This enables us to provide guarantees for the accuracy of any numerical approximation of a given conditional expectation. We illustrate the method by assessing the quality of approximate conditional expectations obtained by linear, polynomial and neural network regression in different concrete examples.

Full work available at URL: https://arxiv.org/abs/2112.01804

Recommendations

• Computing Bounds on Expectations
• Accurate Computation of Conditional Expectation for Highly Nonlinear Problems
• Computing conditional expectations of multidimensional diffusion processes
• scientific article; zbMATH DE number 50805
• Theory & Methods: An Efficient Simulation Method for the Computation of a Class of Conditional Expectations
• On distributions of conditional expectations
• scientific article

zbMATH Keywords

least squares regressionMonte Carlo methodsconditional expectationnumerical guaranteestrustworthy AI

Mathematics Subject Classification ID

General nonlinear regression (62J02) Monte Carlo methods (65C05) Learning and adaptive systems in artificial intelligence (68T05) Probabilistic models, generic numerical methods in probability and statistics (65C20) Error analysis and interval analysis (65G99)

Cites Work

• Title not available (Why is that?)
• Deep learning
• A regression-based Monte Carlo method to solve backward stochastic differential equations
• Regression analysis by example
• Title not available (Why is that?)
• Title not available (Why is that?)
• A probabilistic numerical method for fully nonlinear parabolic PDEs
• Discrete-time approximation and Monte-Carlo simulation of backward stochastic differential equations
• Introduction to stochastic control theory
• Linear regression MDP scheme for discrete backward stochastic differential equations under general conditions
• Nested simulation in portfolio risk measurement
• Computing the distribution function of a conditional expectation via Monte Carlo: discrete conditioning spaces
• Efficient risk estimation via nested sequential simulation
• Title not available (Why is that?)
• Valuing American options by simulation: a simple least-squares approach
• Fundamentals of stochastic filtering
• On the calculation of the solvency capital requirement based on nested simulations
• Stochastic finance. An introduction in discrete time.
• Title not available (Why is that?)
• Valuation of the early-exercise price for options using simulations and nonparametric regression
• Improved lower and upper bound algorithms for pricing American options by simulation
• Risk estimation via regression
• Modern regression methods.
• Meilleures approximations et médianes conditionnelles
• Deep splitting method for parabolic PDEs

Cited In (3)

• Title not available (Why is that?)
• Accurate Computation of Conditional Expectation for Highly Nonlinear Problems
• Computing Bounds on Expectations

This page was built for publication: Computation of conditional expectations with guarantees

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6159022)