Nonlinear and nonnormal filter using importance sampling: antithetic monte carlo integration
- 1 January 1999
- journal article
- research article
- Published by Taylor & Francis in Communications in Statistics - Simulation and Computation
- Vol. 28 (2) , 463-486
- https://doi.org/10.1080/03610919908813560
Abstract
In this paper, the importance sampling filter proposed by Mariano and Tanizaki (1995), Tanizaki (1996), Tanizaki and Mariano (1994) is extended using the antithetic Monte Carlo method to reduce the simulation errors. By Monte Carlo studies, the importance sampling filter with the antithetic Monte Carlo method is compared with the importance sampling filter without the antithetic Monte Carlo method. It is shown that for all the simulation studies the former is clearly superior to the latter especially when number of random draws is smallKeywords
This publication has 16 references indexed in Scilit:
- Estimation of stochastic volatility models via Monte Carlo maximum likelihoodJournal of Econometrics, 1998
- Nonlinear and non-Gaussian state-space modeling with Monte Carlo simulationsJournal of Econometrics, 1998
- Markov chain Monte Carlo in conditionally Gaussian state space modelsBiometrika, 1996
- Markov Chain Monte Carlo Simulation Methods in EconometricsEconometric Theory, 1996
- On Gibbs sampling for state space modelsBiometrika, 1994
- Prediction, filtering and smoothing in non‐linear and non‐normal cases using Monte Carlo integrationJournal of Applied Econometrics, 1994
- RECENT PROGRESS IN APPLIED BAYESIAN ECONOMETRICSJournal of Economic Surveys, 1994
- A Monte Carlo Approach to Nonnormal and Nonlinear State-Space ModelingJournal of the American Statistical Association, 1992
- Bayesian Inference in Econometric Models Using Monte Carlo IntegrationEconometrica, 1989
- Monte Carlo Approximations in Bayesian Decision TheoryJournal of the American Statistical Association, 1989