title: A New Family of Extended Baum-Welch Update Rules 
creator: Kanevsky, Dr Dimitri
creator: Povey, Dr Daniel
creator: Ramabhadran, Dr Bhuvana
creator: Rish, Dr Irina
creator: Sainath, Dr Tara
subject: Speech
description: In this paper, we consider a generalization of the state-of-art  discriminative method for optimizing the conditional likelihood in  Hidden Markov Models (HMMs), called the Extended Baum-Welch (EBW)  algorithm, that has had significant impact on the speech recognition  community. We propose a generalized form of EBW update rules that  can be associated with a weighted sum of updated and initial models,  and  demonstrate that using novel update rules can significantly  speed up parameter estimation for Gaussian mixtures.
date: 2008-04-27
type: Preprint
type: NonPeerReviewed
format: application/pdf
identifier: http://cogprints.org/6037/1/ebw.pdf
identifier:   Kanevsky, Dr Dimitri and Povey, Dr Daniel and Ramabhadran, Dr Bhuvana and Rish, Dr Irina and Sainath, Dr Tara  (2008) A New Family of Extended Baum-Welch Update Rules.  [Preprint]     
relation: http://cogprints.org/6037/