Multilayer perceptrons for state-dependent weightings of HMM likelihoods.
Yong-Joo ChungChong Kwan UnPublished in: Speech Commun. (1996)
Keyphrases
- state dependent
- multilayer perceptron
- hidden markov models
- steady state
- neural network
- artificial neural networks
- queueing networks
- optimal policy
- radial basis function
- neural network model
- back propagation
- stationary distribution
- markov chain
- product form
- arrival rate
- hidden layer
- service rates
- random variables
- queue length
- single server
- maximum likelihood
- asymptotically optimal
- radial basis function network
- activation function
- random walk
- dynamic programming
- data mining
- markov model
- markov models
- feedforward neural networks
- state space
- training data