Marginal likelihood for parallel series

DOI10.3150/07-BEJ119zbMATH Open1157.62058arXiv0810.3978OpenAlexW3105802281MaRDI QIDQ1002541FDOQ1002541

Authors: P. McCullagh 

Publication date: 2 March 2009

Published in: Bernoulli (Search for Journal in Brave)

Abstract: Suppose that $k$ series, all having the same autocorrelation function, are observed in parallel at $n$ points in time or space. From a single series of moderate length, the autocorrelation parameter $$ can be estimated with limited accuracy, so we aim to increase the information by formulating a suitable model for the joint distribution of all series. Three Gaussian models of increasing complexity are considered, two of which assume that the series are independent. This paper studies the rate at which the information for $$ accumulates as $k$ increases, possibly even beyond $n$. The profile log likelihood for the model with $k(k+1)/2$ covariance parameters behaves anomalously in two respects. On the one hand, it is a log likelihood, so the derivatives satisfy the Bartlett identities. On the other hand, the Fisher information for $$ increases to a maximum at $k=n/2$, decreasing to zero for $kgen$. In any parametric statistical model, one expects the Fisher information to increase with additional data; decreasing Fisher information is an anomaly demanding an explanation.

Full work available at URL: https://arxiv.org/abs/0810.3978

Recommendations

• scientific article; zbMATH DE number 4005394
• On the equivalence of marginal and approximate conditional likelihoods for correlation parameters under a normal model
• A time series illustration of approximate conditional likelihood
• Parameter estimation in a stationary autoregressive process with correlated multiple observations
• A note on pseudolikelihood constructed from marginal densities

zbMATH Keywords

profile likelihoodancillary statisticmarginal likelihoodphylogeneticsgroup orbitBartlett identitycombination of informationdecreasing Fisher informationrandom orthogonal matrix

Mathematics Subject Classification ID

Time series, auto-correlation, regression, etc. in statistics (GARCH) (62M10) Non-Markovian processes: estimation (62M09) Problems related to evolution (92D15)

Cites Work

• Interpolation of spatial data. Some theory for kriging
• Title not available (Why is that?)
• Title not available (Why is that?)
• Maximum Likelihood Approaches to Variance Component Estimation and to Related Problems
• Title not available (Why is that?)
• Recovery of inter-block information when block sizes are unequal
• Space–Time Covariance Functions
• Consistent Estimates Based on Partially Consistent Observations
• Title not available (Why is that?)
• A likelihood ratio test for separability of covariances
• The Efficient Generation of Random Orthogonal Matrices with an Application to Condition Estimators
• Title not available (Why is that?)
• A time series illustration of approximate conditional likelihood
• APPROXIMATE CONFIDENCE INTERVALS
• On the Eigenvalues of Random Matrices
• Title not available (Why is that?)
• Algorithm AS 127: Generation of Random Orthogonal Matrices
• Marginal likelihood methods for distributed lag models
• On the equivalence of marginal and approximate conditional likelihoods for correlation parameters under a normal model
• Evidence for conformal invariance of crop yields
• Title not available (Why is that?)

Cited In (1)

• Affine-transformation invariant clustering models

Uses Software

• AS 127

This page was built for publication: Marginal likelihood for parallel series

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1002541)