Statistics – Computation
Scientific paper
2010-04-04
Statistics
Computation
24 pages, 9 figures, and 6 tables
Scientific paper
The ECME algorithm has proven to be an effective way of accelerating the EM algorithm for many problems. Recognising the limitation of using prefixed acceleration subspace in ECME, we propose the new Dynamic ECME (DECME) algorithm which allows the acceleration subspace to be chosen dynamically. Our investigation of an inefficient special case of DECME, the classical Successive Overrelaxation (SOR) method, leads to an efficient, simple, and widely applicable DECME implementation, called DECME_v1. The fast convergence of DECME_v1 is established by the theoretical result that, in a small neighbourhood of the maximum likelihood estimate (MLE), DECME_v1 is equivalent to a conjugate direction method. Numerical results show that DECME_v1 and its two variants are very stable and often converge faster than EM by a factor of one hundred in terms of number of iterations and a factor of thirty in terms of CPU time when EM is very slow.
He Yunxiao
Liu Chuanhai
No associations
LandOfFree
The Dynamic ECME Algorithm does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with The Dynamic ECME Algorithm, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and The Dynamic ECME Algorithm will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-448939