The normal law under linear restrictions: simulation and estimation via minimax tilting

DOI10.1111/RSSB.12162zbMATH Open1414.62102arXiv1603.04166OpenAlexW3106380050WikidataQ29540578 ScholiaQ29540578MaRDI QIDQ5378158FDOQ5378158

Authors: Z. I. Botev, Z. I. Botev 

Publication date: 12 June 2019

Published in: Journal of the Royal Statistical Society Series B: Statistical Methodology (Search for Journal in Brave)

Abstract: Simulation from the truncated multivariate normal distribution in high dimensions is a recurrent problem in statistical computing, and is typically only feasible using approximate MCMC sampling. In this article we propose a minimax tilting method for exact iid simulation from the truncated multivariate normal distribution. The new methodology provides both a method for simulation and an efficient estimator to hitherto intractable Gaussian integrals. We prove that the estimator possesses a rare vanishing relative error asymptotic property. Numerical experiments suggest that the proposed scheme is accurate in a wide range of setups for which competing estimation schemes fail. We give an application to exact iid simulation from the Bayesian posterior of the probit regression model.

Full work available at URL: https://arxiv.org/abs/1603.04166

Recommendations

• Fast simulation of hyperplane-truncated multivariate normal distributions
• Efficient algorithms for generating truncated multivariate normal distributions
• Fast simulation of truncated Gaussian distributions
• Efficient sampling methods for truncated multivariate normal and Student-\(t\) distributions subject to linear inequality constraints
• A New Rejection Sampling Method for Truncated Multivariate Gaussian Random Variables Restricted to Convex Sets

zbMATH Keywords

multivariate normal distributionexponential tiltingexact simulationlinear inequalitiespolytope probabilitiesprobit posterior simulation

Mathematics Subject Classification ID

Monte Carlo methods (65C05) Estimation in multivariate analysis (62H12) Parametric inference under constraints (62F30) Randomized algorithms (68W20)

Cited In (42)

• Bayesian boundary trend filtering
• Approximating Gaussian Process Emulators with Linear Inequality Constraints and Noisy Observations via MC and MCMC
• VeccTMVN
• Constrained interest rates and changing dynamics at the zero lower bound
• Model Calibration With Censored Data
• A Mass-Shifting Phenomenon of Truncated Multivariate Normal Priors
• Distribution-free algorithms for predictive stochastic programming in the presence of streaming data
• Fast simulation of truncated Gaussian distributions
• Collocation Methods for Exploring Perturbations in Linear Stability Analysis
• Gaussian process optimization with failures: classification and convergence proof
• Title not available (Why is that?)
• Tractable Bayes of Skew-Elliptical Link Models for Correlated Binary Data
• Exploiting low-rank covariance structures for computing high-dimensional normal and Student-\(t\) probabilities
• Quasi-Monte Carlo Methods for Binary Event Models with Complex Family Data
• Expectation propagation for the smoothing distribution in dynamic probit
• Forecasting with a panel Tobit model
• Technology choice under emission regulation uncertainty in international container shipping
• Quantitative Verification for Neural Networks using ProbStars
• Co-localization analysis in fluorescence microscopy via maximum entropy copula
• A closed-form filter for binary time series
• Copulas and Histogram-Valued Data
• Bayesian Inference on High-Dimensional Multivariate Binary Responses
• Finite-Dimensional Gaussian Approximation with Linear Inequality Constraints
• Semi-parametric contextual bandits with graph-Laplacian regularization
• Scalable Computation of Predictive Probabilities in Probit Models with Gaussian Process Priors
• Title not available (Why is that?)
• A unified framework for closed-form nonparametric regression, classification, preference and mixed problems with skew Gaussian processes
• Bayesian Conjugacy in Probit, Tobit, Multinomial Probit and Extensions: A Review and New Results
• Fast and accurate computation of the distribution of sums of dependent log-normals
• Skew Gaussian processes for classification
• Large Bayesian SVARs with linear restrictions
• Bayesian-EUCLID: discovering hyperelastic material laws with uncertainties
• Title not available (Why is that?)
• A derivative-free trust-region algorithm with copula-based models for probability maximization problems
• Efficient Bayesian shape-restricted function estimation with constrained Gaussian process priors
• Discrete-time linear skew-Gaussian system and its recursive fixed-dimensional exact density filtering
• Efficient simulation for dependent rare events with applications to extremes
• Generalized adaptive partition-based method for two-stage stochastic linear programs with fixed recourse
• Bayesian analysis of nonparanormal graphical models using rank-likelihood
• Kernel-based identification using Lebesgue-sampled data
• Stochastic epidemic metapopulation models on networks: SIS dynamics and control strategies
• An \(O(N)\) algorithm for computing expectation of \(N\)-dimensional truncated multi-variate normal distribution. I: Fundamentals

This page was built for publication: The normal law under linear restrictions: simulation and estimation via minimax tilting

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5378158)