Combining heterogeneous deep neural networks with conditional random fields for Chinese dialogue act recognition.
Yucan ZhouQinghua HuJie LiuYuan JiaPublished in: Neurocomputing (2015)
Keyphrases
- conditional random fields
- neural network
- graphical models
- hidden markov models
- sequence labeling
- random fields
- probabilistic model
- crf model
- maximum entropy
- supervised training
- information extraction
- markov random field
- object recognition
- higher order
- semi markov
- generative model
- structured learning
- feature extraction
- markov models
- parameter learning
- image labeling
- web page prediction
- named entity recognition
- structured prediction
- superpixels
- recognition process
- protein fold recognition
- error correcting output coding
- segmentation method
- back propagation
- prediction accuracy
- graph cuts
- response time
- pairwise
- feature selection
- information retrieval