Abstract:Importance sampling (IS) is valuable in reducing the variance of Monte Carlo sam-pling for many areas, including finance, rare event simulation, and Bayesian inference. It is natural and obvious to combine quasi-Monte Carlo (QMC) methods with IS to achieve a faster rate of con-vergence. However, a naive replacement of Monte Carlo with QMC may not work well. This paper investigates the convergence rates of randomized QMC-based IS for estimating integrals with respect to a Gaussian measure, in which the IS measure is a Gaussian or t distribution. We prove that if the target function satisfies the so-called boundary growth condition and the covariance matrix of the IS density has eigenvalues no smaller than 1, then randomized QMC with the Gaussian proposal has a root mean squared error of O(N -1+\epsilon) for arbitrarily small \epsilon > 0. Similar results with a t distribution as the proposal are also established. These sufficient conditions help to assess the effectiveness of IS in QMC. For some particular applications, we find that the Laplace IS, a very general approach to approximate the target function by a quadratic Taylor approximation around its mode, has ei-genvalues smaller than 1, making the resulting integrand less favorable for QMC. From this point of view, when using Gaussian distributions as the IS proposal, a change of measure via Laplace IS may transform a favorable integrand into an unfavorable one for QMC although the variance of Monte Carlo sampling is reduced. We also study the effect of positivization tricks on the error rate when the integrand has mixed sign. If the smooth positivization proposed by Owen and Zhou (J. Amer. Statist. Assoc., 95 (2000), pp. 135--143) is used, the rate O(N -1+\epsilon) is retained. This is not the case if taking the positive and negative parts of the integrand. We also give some examples to verify our propositions and warn against naive replacement of Monte Carlo with QMC under IS proposals. Numerical results suggest that using Laplace IS with t distributions is more robust than that with Gaussian distributions.

On the empirical approximation to quantiles from Lugannani-Rice saddlepoint formula

The saddlepoint approximation for averages of conditionally independent random variables

Nonparametric Estimation of Quantiles for A Class of Stationary Processes

Saddlepoint Approximations for the Difference of Order Statistics and Studentized Sample Quantiles

Order Estimates for the Exact Lugannani-Rice Expansion

Saddlepoint approximations for likelihood ratio like statistics with applications to permutation tests

An Empirical Quantile Estimation Approach for Chance-Constrained Nonlinear Optimization Problems

A quantile-based probabilistic mean value theorem

Inter-order relations between equivalence for L-quantiles of the Student's t distribution

An Empirical Quantile Estimation Approach to Nonlinear Optimization Problems with Chance Constraints

Quantile function expansion using regularly varying functions

Bahadur Representation of Linear Kernel Quantile Estimator for Stationary Processes

Saddlepoint Approximation for Sample and Bootstrap Quantiles

On the Impact of Approximation Errors on Extreme Quantile Estimation with Applications to Functional Data Analysis

Partial functional linear quantile regression

A New Family of Error Distributions for Bayesian Quantile Regression

Empirical saddlepoint approximations of the Studentized mean under stratified random sampling

On Some Asymptotic Properties and an Almost Sure Approximation of the Normalized Inverse-Gaussian Process

ON THE ERROR RATE OF IMPORTANCE SAMPLING WITH RANDOMIZED QUASI-MONTE CARLO*

Nearly Semiparametric Efficient Estimation of Quantile Regression

Empirical likelihood for quantile regression models with longitudinal data