EM's monotonicity and local rate are unified by the spectral operator G = I - DT that equals the missing-information ratio and observed-likelihood Hessian, enabling accelerated local updates.
Maximum likelihood from incomplete data via the EM algorithm
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Expectation-Maximization as a Spectrally Governed Relaxation Flow
EM's monotonicity and local rate are unified by the spectral operator G = I - DT that equals the missing-information ratio and observed-likelihood Hessian, enabling accelerated local updates.