CHEMDNER system with mixed conditional random fields and multi-scale word clustering.
Yanan LuDonghong JiXiaoyuan YaoXiaomei WeiXiaohui LiangPublished in: J. Cheminformatics (2015)
Keyphrases
- conditional random fields
- web page prediction
- multiscale
- graphical models
- random fields
- clustering algorithm
- hidden markov models
- sequence labeling
- probabilistic model
- information extraction
- maximum entropy
- markov models
- higher order
- prediction accuracy
- crf model
- markov random field
- k means
- generative model
- named entity recognition
- error correcting output coding
- co occurrence
- pairwise
- image labeling
- superpixels
- natural images
- structured prediction
- protein fold recognition
- prefetching
- approximate inference
- markov networks
- parameter learning
- contextual features
- decision trees
- unsupervised learning
- keywords
- semi markov
- structured output
- structured learning
- data points
- semi supervised