BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:BSU Seminar: &quot\;Bayesian nonparametric spectral analysis of mu
 ltivariate time series&quot\; - Renate Meyer\, University of Aukland
DTSTART:20240827T130000Z
DTEND:20240827T140000Z
UID:TALK219757@talks.cam.ac.uk
CONTACT:Alison Quenault
DESCRIPTION:The analysis of multivariate time series can give important in
 sights into periodicities and coherencies. We present a  novel approach to
  Bayesian nonparametric spectral analysis of stationary multivariate time 
 series which is based on Whittle's likelihood. Starting with a parametric\
 nvector-autoregressive model\, the parametric likelihood is nonparametrica
 lly adjusted in the frequency domain to account for potential deviations f
 rom parametric assumptions. We show contiguity of the nonparametrically co
 rrected likelihood\, the multivariate Whittle likelihood approximation and
  the exact likelihood for Gaussian time series. The nonparametric prior us
 ed is a multivariate extension of the nonparametric Bernstein-Dirichlet pr
 ocess prior for univariate spectral\ndensities to the space of Hermitian p
 ositive definite spectral density matrices. An infinite series representat
 ion of this prior is then used to develop a Markov chain Monte Carlo algor
 ithm to sample from the posterior distribution. The code is made publicly 
 available for ease of use and reproducibility. With this novel approach we
  provide a generalization of the multivariate Whittle-likelihood-based met
 hod of Meier et al. (2020) as well as an extension of the nonparametricall
 y corrected likelihood for univariate stationary time series of Kirch et a
 l. (2019) to the multivariate case. In a simulation study\, we demonstrate
  that the nonparametrically corrected likelihood combines the efficiencies
  of a parametric with the robustness of a nonparametric model. We illustra
 te the practical benefits using case studies of EEG and windspeed time ser
 ies.
LOCATION:MRC Biostatistics Unit\, East Forvie Building\, Forvie Site Robin
 son Way Cambridge CB2 0SR.
END:VEVENT
END:VCALENDAR