Variational Bayes with synthetic likelihood
From MaRDI portal
Publication:1704030
Abstract: Synthetic likelihood is an attractive approach to likelihood-free inference when an approximately Gaussian summary statistic for the data, informative for inference about the parameters, is available. The synthetic likelihood method derives an approximate likelihood function from a plug-in normal density estimate for the summary statistic, with plug-in mean and covariance matrix obtained by Monte Carlo simulation from the model. In this article, we develop alternatives to Markov chain Monte Carlo implementations of Bayesian synthetic likelihoods with reduced computational overheads. Our approach uses stochastic gradient variational inference methods for posterior approximation in the synthetic likelihood context, employing unbiased estimates of the log likelihood. We compare the new method with a related likelihood free variational inference technique in the literature, while at the same time improving the implementation of that approach in a number of ways. These new algorithms are feasible to implement in situations which are challenging for conventional approximate Bayesian computation (ABC) methods, in terms of the dimensionality of the parameter and summary statistic.
Recommendations
Cites work
- A comparative review of dimension reduction methods in approximate Bayesian computation
- A comparison of inferential methods for highly nonlinear state space models in ecology and epidemiology
- A Stochastic Approximation Method
- An extended empirical saddlepoint approximation for intractable likelihoods
- Approximate Bayesian computational methods
- Automatic differentiation variational inference
- Bayesian model comparison with un-normalised likelihoods
- Bayesian optimization for likelihood-free inference of simulator-based statistical models
- Bayesian Synthetic Likelihood
- Efficient implementation of Markov chain Monte Carlo when using an unbiased likelihood estimator
- Expectation propagation for likelihood-free inference
- Explaining variational approximations
- Extending approximate Bayesian computation methods to high dimensions via a Gaussian copula model
- Fixed-form variational posterior approximation through stochastic linear regression
- Fully simplified multivariate normal updates in non-conjugate variational message passing
- Gaussian variational approximation with sparse precision matrices
- scientific article; zbMATH DE number 6377992 (Why is no real title available?)
- scientific article; zbMATH DE number 1221798 (Why is no real title available?)
- scientific article; zbMATH DE number 1261669 (Why is no real title available?)
- scientific article; zbMATH DE number 1134711 (Why is no real title available?)
- scientific article; zbMATH DE number 947414 (Why is no real title available?)
- scientific article; zbMATH DE number 835699 (Why is no real title available?)
- scientific article; zbMATH DE number 3047450 (Why is no real title available?)
- Large-scale machine learning with stochastic gradient descent
- Likelihood-free Bayesian estimation of multivariate quantile distributions
- Likelihood-free Bayesian inference for \(\alpha\)-stable models
- Likelihood-free inference via classification
- On some properties of Markov chain Monte Carlo simulation methods based on the particle filter
- Pattern recognition and machine learning.
- Pre-processing for approximate Bayesian computation in image analysis
- PRMLT
- Sampling and Bayes' Inference in Scientific Modelling and Robustness
- Simple consistent estimators of stable distribution parameters
- The pseudo-marginal approach for efficient Monte Carlo computations
- Unbiased Estimation of Some Multivariate Probability Densities and Related Functions
- Weighted quantile-based estimation for a class of transformation distributions.
Cited in
(21)- An approximate likelihood perspective on ABC methods
- Likelihood-free inference in high dimensions with synthetic likelihood
- Variational Bayes on manifolds
- Likelihood-free inference with deep Gaussian processes
- Fast and accurate variational inference for models with many latent variables
- Likelihood-free approximate Gibbs sampling
- Multifidelity multilevel Monte Carlo to accelerate approximate Bayesian parameter inference for partially observed stochastic processes
- scientific article; zbMATH DE number 6377992 (Why is no real title available?)
- Variational Bayes Estimation of Discrete-Margined Copula Models With Application to Time Series
- Bayesian Deep Net GLM and GLMM
- Variational Bayesian Filtering
- Efficient Bayesian Synthetic Likelihood With Whitening Transformations
- Unbiased MLMC-based variational Bayes for likelihood-free inference
- Bayesian Inference Using Synthetic Likelihood: Asymptotics and Adjustments
- Stochastic variational inference for GARCH models
- Approximating Bayes in the 21st century
- Analytic natural gradient updates for Cholesky factor in Gaussian variational approximation
- Natural Gradient Variational Bayes Without Fisher Matrix Analytic Calculation and Its Inversion
- Wasserstein Gaussianization and Efficient Variational Bayes for Robust Bayesian Synthetic Likelihood
- Bayesian Synthetic Likelihood
- Robust Bayesian synthetic likelihood via a semi-parametric approach
This page was built for publication: Variational Bayes with synthetic likelihood
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1704030)
