COMPETITIVE TRAINING - A CONNECTIONIST APPROACH TO THE DISCRIMINATIVE TRAINING OF HIDDEN MARKOV-MODELS

The paper presents hidden Markov models (HMMs) within a connectionist framework and shows how error back propagation can be used to discriminatively train HMM parameters. The relationship between this competitive training approach and conventional Baum–Welch re-estimation is explored and experimental results presented for its application in ergodic HMM architectures.