The Hyvärinen scoring rule in Gaussian linear time series models

DOI10.1016/J.JSPI.2020.08.004zbMATH Open1465.62147OpenAlexW3095535930MaRDI QIDQ830689FDOQ830689

Authors Silvia Columbu, Valentina Mameli, Monica Musio, Philip Dawid

Publication date 7 May 2021

Published in Journal of Statistical Planning and Inference (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1904.12513

zbMATH Keywords

Gaussian linear time seriesscoring rule estimatorsHyvärinen scoring rule

Mathematics Subject Classification ID

Density estimation (62G07) Linear regression; mixed models (62J05) Time series, auto-correlation, regression, etc. in statistics (GARCH) (62M10) Numerical analysis or methods applied to Markov chains (65C40) Gaussian processes (60G15) White noise theory (60H40)

Abstract: Likelihood-based estimation methods involve the normalising constant of the model distributions, expressed as a function of the parameter. However in many problems this function is not easily available, and then less efficient but more easily computed estimators may be attractive. In this work we study stationary time-series models, and construct and analyse "score-matching estimators, that do not involve the normalising constant. We consider two scenarios: a single series of increasing length, and an increasing number of independent series of fixed length. In the latter case there are two variants, one based on the full data, and another based on a sufficient statistic. We study the empirical performance of these estimators in three special cases, autoregressive (AR), moving average (MA) and fractionally differenced white noise (ARFIMA) models, and make comparisons with full and pairwise likelihood estimators. The results are somewhat model-dependent, with the new estimators doing well for $MA$ and ARFIMA models, but less so for $AR$ models.

Recommendations

• Accelerating score-driven time series models
• Linear estimating equations for exponential families with application to Gaussian linear concentration models
• Bayesian model comparison with the Hyvärinen score: computation and consistency
• On a criterion for the selection of models for stationary time series
• Some extensions of score matching

Cites work

• scientific article; zbMATH DE number 5849508 (Why is no real title available?)
• scientific article; zbMATH DE number 3565994 (Why is no real title available?)
• scientific article; zbMATH DE number 735225 (Why is no real title available?)
• scientific article; zbMATH DE number 1122119 (Why is no real title available?)
• A note on Bayesian model selection for discrete data using proper scoring rules
• Bayesian model comparison with the Hyvärinen score: computation and consistency
• Bayesian model selection based on proper scoring rules
• Comments on pairwise likelihood in time series models
• Empirical and simulated adjustments of composite likelihood ratio statistics
• Estimation of non-normalized statistical models by score matching
• Fractional differencing
• Inverse Autocorrelations
• Local proper scoring rules of order two
• Minimum scoring rule inference
• On Moments of the Inverted Wishart Distribution
• On composite marginal likelihoods
• On weighting of bivariate margins in pairwise likelihood
• Proper local scoring rules
• Proper local scoring rules on discrete sample spaces
• Scoring rules for statistical models on spheres
• The Inverse Autocorrelations of a Time Series and Their Applications
• The asymptotic distribution of the sample inverse autocorrelations of an autoregressive-moving average process
• The geometry of proper scoring rules
• Theory and applications of proper scoring rules
• Time series: theory and methods.

Describes a project that uses

Uses Software

• R
• numDeriv

This page was built for publication: The Hyvärinen scoring rule in Gaussian linear time series models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q830689)