A dual-homotopy framework enables a constrained EM algorithm that preserves likelihood monotonicity and produces more stable estimates than standard or existing constrained EM methods.
Statistics and Decisions, Dedewicz , volume=
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
-
Dual-Homotopy Framework for Constrained EM Algorithm
A dual-homotopy framework enables a constrained EM algorithm that preserves likelihood monotonicity and produces more stable estimates than standard or existing constrained EM methods.