147 search results - page 22 / 30 » Neural Network Classification and Prior Class Probabilities |
110
click to vote
IWANN
2005
Springer
15 years 8 months ago
2005
Springer
A large class of problems requires real-time processing of complex temporal inputs in real-time. These are difficult tasks for state-of-the-art techniques, since they require captu...
148
click to vote
ICML
2004
IEEE
16 years 4 months ago
2004
IEEE
Bayesian networks are a powerful probabilistic representation, and their use for classification has received considerable attention. However, they tend to perform poorly when lear...
118
click to vote
ICASSP
2008
IEEE
15 years 9 months ago
2008
IEEE
In this paper, we investigate the significance of contextual information in a phoneme recognition system using the hidden Markov model - artificial neural network paradigm. Cont...
130
click to vote
TIT
1998
15 years 2 months ago
1998
—Probability models are estimated by use of penalized log-likelihood criteria related to AIC and MDL. The accuracies of the density estimators are shown to be related to the trad...
129
click to vote
COLING
2002
15 years 3 months ago
2002
Language users have individual linguistic styles. A spoken dialogue system may benefit from adapting to the linguistic style of a user in input analysis and output generation. To ...