Title :
On Convergence Properties of Message-Passing Estimation Algorithms
Author_Institution :
RIKEN Brain Sci. Inst., Saitama
Abstract :
The convergence and stability properties of several message-passing estimation algorithms are investigated, i.e., (generalized) expectation maximization, gradient methods, and coordinate ascent/descent. Results are presented for cycle-free and cyclic factor graphs.
Keywords :
estimation theory; expectation-maximisation algorithm; gradient methods; graph theory; message passing; signal processing; convergence property; coordinate ascent/descent algorithm; cycle-free graph; cyclic factor graph; generalized expectation maximization; gradient method; message-passing estimation algorithm; stability property; Convergence; Gradient methods; Heart; Hidden Markov models; Iterative algorithms; Message passing; Parameter estimation; Stability; State-space methods; Stochastic systems;
Conference_Titel :
Information Theory, 2007. ISIT 2007. IEEE International Symposium on
Conference_Location :
Nice
Print_ISBN :
978-1-4244-1397-3
DOI :
10.1109/ISIT.2007.4557604
