Laplace approximation and natural gradient for Gaussian process regression with heteroscedastic Student-\(t\) model

DOI10.1007/s11222-018-9836-0zbMath1430.62046arXiv1712.07437OpenAlexW2963161900MaRDI QIDQ2329797

Marcelo Hartmann, Jarno P Vanhatalo

Publication date: 18 October 2019

Published in: Statistics and Computing (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/1712.07437

zbMATH Keywords

Riemannian metricGaussian processesFisher information matrixLaplace approximationheteroscedastic noiseapproximate inferencelocation-scale regressionnatural gradientStudent-\(t\) model

Mathematics Subject Classification ID

Gaussian processes (60G15) Statistics on manifolds (62R30) Parametric hypothesis testing (62F03) General nonlinear regression (62J02) Monte Carlo methods (65C05) Statistical aspects of information-theoretic topics (62B10)

Related Items (3)

Identification of Gaussian process with switching noise mode and missing data ⋮ Estimation and prediction of a generalized mixed-effects model with \(t\)-process for longitudinal correlated binary data ⋮ Mixture of multivariate Gaussian processes for classification of irregularly sampled satellite image time-series

Uses Software

• GPstuff
• GPML
• PRMLT

Cites Work

• Understanding predictive information criteria for Bayesian models
• Posterior property of Student-\(t\) linear regression model using objective priors
• Newton's method and its use in optimization
• Multivariate adaptive regression splines
• Hedonic housing prices and the demand for clean air
• Choice of basis for Laplace approximation
• Some diagnostics of maximum likelihood and posterior nonnormality
• Robust diagnostic regression analysis
• Penalising model component complexity: a principled, practical approach to constructing priors
• A survey of Bayesian predictive methods for model assessment, selection and comparison
• On the comparison of Fisher information of the Weibull and GE distributions
• Approximate Bayesian Inference for Latent Gaussian models by using Integrated Nested Laplace Approximations
• Posterior expectations for large observations
• Objective Bayesian analysis for the Student-t regression model
• Accurate Approximations for Posterior Moments and Marginal Densities
• Approximate Bayes factors and accounting for model uncertainty in generalised linear models
• Newton-Raphson and Related Algorithms for Maximum Likelihood Variance Component Estimation
• Multivariate Student-t regression models: Pitfalls and inference
• Riemann Manifold Langevin and Hamiltonian Monte Carlo Methods
• Fully Exponential Laplace Approximations to Expectations and Variances of Nonpositive Functions
• Bayes Factors
• THE PROBABLE ERROR OF A MEAN
• Information-Geometric Optimization Algorithms: A Unifying Picture via Invariance Principles
• Prior distributions for variance parameters in hierarchical models (Comment on article by Browne and Draper)
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item

This page was built for publication: Laplace approximation and natural gradient for Gaussian process regression with heteroscedastic Student-\(t\) model