Low-Rank Hidden State Embeddings for Viterbi Sequence Labeling.
Dung ThaiShikhar MurtyTrapit BansalLuke VilnisDavid BelangerAndrew McCallumPublished in: CoRR (2017)
Keyphrases
- low rank
- markov models
- hidden markov models
- conditional random fields
- high dimensional data
- missing data
- linear combination
- convex optimization
- markov model
- maximum entropy
- matrix factorization
- higher order
- dimensionality reduction
- low dimensional
- semi supervised
- high order
- structured prediction
- generative model
- information extraction
- probabilistic model
- nearest neighbor
- markov networks
- pairwise
- missing values
- machine learning
- maximum likelihood
- graphical models
- high dimensional
- approximate inference
- support vector
- information retrieval