Feature Noising for Log-Linear Structured Prediction.
Sida I. WangMengqiu WangStefan WagerPercy LiangChristopher D. ManningPublished in: EMNLP (2013)
Keyphrases
- structured prediction
- log linear
- max margin
- maximum margin
- latent variables
- efficient learning
- log linear models
- conditional random fields
- markov networks
- convex optimization
- support vector
- discriminative training
- pattern classification
- approximate inference
- belief networks
- feature set
- feature selection
- random variables
- graphical models
- multi class
- probabilistic model
- prior knowledge
- feature vectors
- feature extraction