Maximum likelihood estimation using composite likelihoods for closed exponential families

By Kanti V. Mardia, John T. Kent, Gareth Hughes, Charles C. Taylor. 2009

69 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/d446ad963b54ec0d3aed3b4f315f091f2680b3dc

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistics and Probability

Citations in the corpus

The Composite Marginal Likelihood (CML) Inference Approach with Applications to Discrete and Mixed Dependent Variable Models (2014, methodology): “Examples of studies that have used precise analytic computations to compare the asymptotic efficiency of the ML and CML estimators include Cox and Reid (2004), Hjort and Varin (2008), and Mardia et al. (2009).”
AN OVERVIEW OF COMPOSITE LIKELIHOOD METHODS (2011, methodology): “Mardia et al. (2009) provide an explanation for this, by showing that composite conditional estimators are fully efficient in exponential families that have a certain closure property under subsetting.”
COMMENTS ON PAIRWISE LIKELIHOOD IN TIME SERIES MODELS (2011, background): “In some cases it is even fully efficient, as seen in the AR(1) example in Section 3; see also Mardia et al. (2009).”
ADJUSTING COMPOSITE LIKELIHOOD RATIO STATISTICS (2009, background): “(4.2) This model is rather peculiar in the sense that the estimate from the pairwise likelihood coincides with full maximum likelihood estimate, θ̂p = θ̂ (Mardia et al. (2009)).”
Fitting Vast Dimensional Time-Varying Covariance Models (2020, unknown): “The exact magnitude of this loss will depend on the model under consideration (see, e.g., Mardia et al. 2009; Kenne Pagui, Salvan, and Sartori 2015).”
The maximum approximate composite marginal likelihood (MACML) estimation of multinomial probit-based unordered response choice models (2011, result): “7 A handful of studies (see Hjort and Varin, 2008; Mardia et al., 2009; Cox and Reid, 2004) have also theoretically examined the limiting normality properties of the CML approach, and compared the asymptotic variance matrices from this approach and the ML approach.”
A comparison of the maximum simulated likelihood and composite marginal likelihood estimation approaches in the context of the multivariate ordered-response model (2010, background): “In the low correlation case, the absolute percentage bias (APB) ranges from (10) McFadden and Train (2000) indicate, in their use of independent number of random draws across observations, that the difference between the asymptotic covariance matrix of the MSL estimator obtained as the inverse of th”