Sciweavers

CSDA
2007
94views more  CSDA 2007»
13 years 3 months ago
Some extensions of score matching
Many probabilistic models are only defined up to a normalization constant. This makes maximum likelihood estimation of the model parameters very difficult. Typically, one then h...
Aapo Hyvärinen
NAACL
2004
13 years 5 months ago
Training Tree Transducers
Many probabilistic models for natural language are now written in terms of hierarchical tree structure. Tree-based modeling still lacks many of the standard tools taken for grante...
Jonathan Graehl, Kevin Knight