Inference for clustered data using the independence loglikelihood

By Richard E. Chandler, Steven Bate. 2007

209 citations recorded by the source corpus.

Subfields: Finance, Global and Planetary Change

View Paper: https://www.semanticscholar.org/paper/eb3ca54a46327c185498174f921869472dd65f0b

How researchers see this paper

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

In Finance

In Global and Planetary Change

Citations in the corpus

Non‐stationary intensity‐duration‐frequency curves integrating information concerning teleconnections and climate change (2018, background): “This function is often referred to as the independence likelihood (Chandler and Bate, 2007) and is given by:”
Geostatistics of extremes (2012, methodology): “In a similar situation, Chandler & Bate (2007) suggest modifications which ensure that likelihood ratio statistics retain their usual large sample distributions.”
Bayesian Inference from Composite Likelihoods, with an Application to Spatial Extremes (2009, methodology): “One possible choice, C = MMA, where M T AMA = H(θ0)J(θ0) H(θ0) and MTM = H(θ0), corresponds to a suggestion of Chandler and Bate (2007) for hypothesis testing for clustered data using the independence log-likelihood.”
Statistical Properties of Covariance Tapers (2013, methodology): “Inference based on approximated likelihoods that goes beyond using the Godambe information matrix is addressed in, for example, Chandler and Bate (2007) and Pace, Salvan, and Sartori (2011).”
AN OVERVIEW OF COMPOSITE LIKELIHOOD METHODS (2011, background): “Chandler and Bate (2007) proposed a different type of adjustment for the independence likelihood: essentially stretching the composite log-likelihood on the θ-axis, about θ̂CL to ensure, at least approximately, that the second Bartlett identity holds, and thus that the usual χ(2)q approximation can ”
ADJUSTING COMPOSITE LIKELIHOOD RATIO STATISTICS (2009, background): “Chandler and Bate (2007) propose the vertical scaling of cw(θ), cw(θ)CB = (θ̂c − θ)>G(θ̂c)(θ̂c − θ) (θ̂c − θ)>H(θ̂c)(θ̂c − θ) cw(θ) , (3.4) having the usual χ2d asymptotic null distribution.”
Bayesian composite marginal likelihoods (2011, methodology): “There are adjustments based on moment matching conditions (Geys, Molenberghs and Ryan (1999)), there is a Satterthwaite-type adjustment suggested in Varin (2008), and there are other proposals based on suitable vertical scalings of W c (θ) (Chandler and Bate (2007)).”
Efficient inference and simulation for elliptical Pareto processes (2013, methodology): “Using threshold exceedances over marginal thresholds taken to be the 95% percentiles at each of the 25 stations, we estimated parameters by maximizing the independence likelihood (Chandler & Bate, 2007), in which the data at each location are wrongly assumed to be independent.”
Bayesian estimation of rainfall intensity-duration-frequency relationships (2015, methodology): “The term independence (log-) likelihood was taken from Chandler and Bates (2007), where it was used for dependent data structures in general.”
Modelling Time Series Extremes (2012, methodology): “…to model parametrically, however, and if the main goal is to model non-stationarity, it will be simpler to use an independence likelihood, which treats extreme observations as if they were independent, but then inflates standard errors to allow for unmodelled dependence (Chandler & Bate, 2007).”
A regional peaks‐over‐threshold model in a nonstationary climate (2012, methodology): “To test this hypothesis, we use the independent likelihood ratio statistic, which is a special case of a composite likelihood ratio (CLR) statistic [Chandler and Bate, 2007; Varin et al., 2011]:”
Spatial extremes: Max-stable processes at work (2013, methodology): “Although the inference procedure introduced in Section 5 is non standard, model selection using (adapted) information criterion Varin and Vidoni (2005) or (composite) likelihood ratio tests Chandler and Bate (2007) is possible.”
Extremes on river networks (2015, methodology): “Inference for the extremes of univariate data is well developed [Coles (2001), de Haan and Ferreira (2006), Embrechts, Klüppelberg and Mikosch (1997)], so we merely sketch it in Section 4.2. Statistical inference for multivariate or spatial models is more difficult, as their distributions are rarely”