Introduces Gaussian chain graph models for time series that induce a group sparse plus group low-rank decomposition of inverse spectral density matrices in the frequency domain for identifiability, along with a three-stage regularized estimation procedure shown to be consistent.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ME 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Time Series Gaussian Chain Graph Models
Introduces Gaussian chain graph models for time series that induce a group sparse plus group low-rank decomposition of inverse spectral density matrices in the frequency domain for identifiability, along with a three-stage regularized estimation procedure shown to be consistent.