Dec 22, 2010

likelihood for Generative and discriminative models

Labels:
Generative model
w = \arg \max \sum_n \log p(t_n)p(x_n|t_n,w)

discriminative model
w = \arg \max \sum_n \log p(t_n|x_n,w)

0 comments:

Post a Comment

Subscribe to: Post Comments (Atom)