Reconciling the Gaussian and Whittle likelihood with an application to estimation in the frequency domain
From MaRDI portal
Publication:2054529
DOI10.1214/21-AOS2055zbMATH Open1486.62243arXiv2001.06966OpenAlexW3212452912MaRDI QIDQ2054529FDOQ2054529
Authors: Suhasini Subba Rao, Junho Yang 
Publication date: 3 December 2021
Published in: The Annals of Statistics (Search for Journal in Brave)
Abstract: In time series analysis there is an apparent dichotomy between time and frequency domain methods. The aim of this paper is to draw connections between frequency and time domain methods. Our focus will be on reconciling the Gaussian likelihood and the Whittle likelihood. We derive an exact, interpretable, bound between the Gaussian and Whittle likelihood of a second order stationary time series. The derivation is based on obtaining the transformation which is biorthogonal to the discrete Fourier transform of the time series. Such a transformation yields a new decomposition for the inverse of a Toeplitz matrix and enables the representation of the Gaussian likelihood within the frequency domain. We show that the difference between the Gaussian and Whittle likelihood is due to the omission of the best linear predictions outside the domain of observation in the periodogram associated with the Whittle likelihood. Based on this result, we obtain an approximation for the difference between the Gaussian and Whittle likelihoods in terms of the best fitting, finite order autoregressive parameters. These approximations are used to define two new frequency domain quasi-likelihoods criteria. We show that these new criteria can yield a better approximation of the spectral divergence criterion, as compared to both the Gaussian and Whittle likelihoods. In simulations, we show that the proposed estimators have satisfactory finite sample properties.
Full work available at URL: https://arxiv.org/abs/2001.06966
Recommendations
periodogramdiscrete Fourier transformsecond order stationary time seriesToeplitz inversebiorthogonal transformsquasi-likelihoods
Point estimation (62F10) Time series, auto-correlation, regression, etc. in statistics (GARCH) (62M10) Inference from stochastic processes and spectral analysis (62M15)
Cites Work
- Predictive, finite-sample model choice for time series under stationarity and non-stationarity
- Beyond Whittle: nonparametric correction of a parametric likelihood with a focus on Bayesian time series analysis
- Edge effects and efficient parameter estimation for stationary random fields
- Title not available (Why is that?)
- On the range of validity of the autoregressive sieve bootstrap
- Title not available (Why is that?)
- Fourier analysis of stationary time series in function space
- Bayesian Estimation of the Spectral Density of a Time Series
- Large-sample properties of parameter estimates for strongly dependent stationary Gaussian time series
- Asymptotically efficient autoregressive model selection for multistep prediction
- Gaussian semiparametric estimation of long range dependence
- An efficient taper for potentially overdifferenced long-memory time series
- Local Whittle estimation of fractional integration for nonlinear processes
- Nonstationarity-extended local Whittle estimation
- Large sample inference for long memory processes
- Baxter's inequality for finite predictor coefficients of multivariate long-memory stationary processes
- Explicit representation of finite predictor coefficients and its applications
- Title not available (Why is that?)
- WHITTLE ESTIMATION OF ARCH MODELS
- APPROXIMATE CONFIDENCE INTERVALS
- Time series. Data analysis and theory.
- A necessary and sufficient condition for asymptotic independence of discrete Fourier transforms under short- and long-range dependence
- Order selection for same-realization predictions in autoregressive processes
- Title not available (Why is that?)
- Baxter's inequality and convergence of finite predictors of multivariate stochastic processes
- SPECTRAL ANALYSIS WITH TAPERED DATA
- On the inverses of some patterned matrices arising in the theory of stationary time series
- On the Inversion of the Sample Covariance Matrix in a Stationary Autoregressive Process
- Asymptotic properties of least-squares estimates of parameters of the spectrum of a stationary non-deterministic time-series
- An Asymptotic Result for the Finite Predictor.
- Title not available (Why is that?)
- Small sample effects in time series analysis: A new asymptotic theory and a new estimate
- Nonparametric high resolution spectral estimation
- An approximate inverse for the covariance matrix of moving average and autoregressive processes
- Approximations for stationary covariance matrices and their inverses with application to ARMA models
- The evaluation of certain quadratic forms occurring in autoregressive model fitting
- Duals of random vectors and processes with applications to prediction problems with missing values
- REDUCTION OF THE ASYMPTOTIC BIAS OF AUTOREGRESSIVE AND SPECTRAL ESTIMATORS BY TAPERING
- Orthogonal samples for estimators in time series
- A note on Herglotz's theorem for time series on function spaces
- Title not available (Why is that?)
- The debiased Whittle likelihood
- Spectral methods for small sample time series: A complete periodogram approach
Cited In (14)
- Frequency domain generalized empirical likelihood method
- Higher‐order asymptotics of minimax estimators for time series
- Contiguity of the Whittle measure for a Gaussian time series
- Explicit formulas for the inverses of Toeplitz matrices, with applications
- Spectral methods for small sample time series: A complete periodogram approach
- On the asymptotic behavior of a finite section of the optimal causal filter
- Posterior consistency for the spectral density of non‐Gaussian stationary time series
- Statistical analysis of irregularly spaced spatial data in frequency domain
- On the equivalence of time and frequency domain maximum likelihood estimation
- A prediction perspective on the Wiener–Hopf equations for time series
- A nonparametrically corrected likelihood for Bayesian spectral analysis of multivariate time series
- Inference with the Whittle Likelihood: A Tractable Approach Using Estimating Functions
- Title not available (Why is that?)
- A Note on Whittle's Likelihood
Uses Software
This page was built for publication: Reconciling the Gaussian and Whittle likelihood with an application to estimation in the frequency domain
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2054529)
