2018 10-708 (CMU) Probabilistic Graphical Models {Lecture 10} [HHM and CRF]

标签：use   models   ati   com   combine   connect   inf   http   ffline   

between tags and words, there‘s table 1.

between tags, there‘s table 2.

combine the two tables, p(...) to get the results. 

MRF: factors of the tables not necessarily probabilities

BN: must be probabilities. => BN is easier to learn than MRF

 Maximum-Entropy Markov Model (MEMM)

Marginals:

1) forward: 

 2) Belief:

HMM is generative, modeling joint probability P(x,y)

but tagging just needs  P(y|x)

https://cedar.buffalo.edu/~srihari/CSE574/Discriminative-Generative.pdf

Full obervation!

(like the offline SLAM?)

biased! because we only look at local observation.

P（x_2|x_1） can be called \Psi(x_1,x_2)

If Y_1 ~~~~ Y_{n-2} are connected somehow, what should be changed?

How close the model is closed to the truth.

2018 10-708 (CMU) Probabilistic Graphical Models {Lecture 10} [HHM and CRF]

标签：use   models   ati   com   combine   connect   inf   http   ffline   

原文地址：https://www.cnblogs.com/ecoflex/p/10231319.html