Markov Network Article Index for
Markov 		Website Links For
Markov 		 

Information About

Markov Network
  CATEGORIES ABOUT MARKOV NETWORK
   
artificial intelligence
   
theoretical computer science
   
probability theory
   
networks
   
APPAREL
BABY
BEAUTY
BOOKS
CAR TOYS
CELL PHONES
DVD'S
ELECTRONICS
GOURMET FOOD
GROCERIES
HEALTH & PERSONAL
HOME & GARDEN
JEWELRY
MUSIC
MUSIC INSTRUMENTS
OFFICE PRODUCTS
SOFTWARE
SPORTING GOODS
TOOLS & HARDWARE
TOYS
VIDEO GAMES
SHOPPING HOME
MORE SHOPPING...



Formally, a Markov network consists of:

  • an Undirected Graph ''G'' = (''V'',''E''), where each vertex ''v'' ∈''V'' represents a random variable in \mathcal{X} and each edge {''u'',''v''} ∈ ''E'' represents a dependency between the random variables ''u'' and ''v'',

  • a set of Potential Function s \phi_k (also called ''factors'' or ''clique potentials''), one for each Maximal Clique ''k'' in ''G''. Each \phi_k is a mapping from possible joint assignments (to the elements of ''k'') to Non-negative Real Number s.


The joint distribution represented by a Markov network is given by:

P(X=x) = rac{1}{Z} \prod_{k} \phi_k (x_{ \{ k \}})

where x_{ \{ k \}} is the state of the random variables in the ''k''th clique, and the Normalizing Constant Z (also called a Partition Function ), where

Z = \sum_{x \isin \mathcal{X}} \prod_{k} \phi_k(x_{ \{ k \} }).

In practice, a Markov network is often conveniently expressed as a Log-linear Model , given by

P(X=x) = rac{1}{Z} \exp(\sum_{k} \phi_k (x_{ \{ k \}}))

with normalizing constant Z = \sum_{x \isin \mathcal{X}} \exp (\sum_{k} \phi_k(x_{ \{ k \} })).

The Markov Blanket of a node v_i in a Markov network is defined to be every node with an edge to v_i , i.e. all v_j such that \lbrace v_i, v_j brace \in E. Every node v in a Markov network is Conditionally Independent of every other node given the Markov blanket of v.

As in a Bayesian network, one may calculate the Conditional Distribution of a set of nodes V' = \{ v_1 ,..., v_i \} given values to another set of nodes W' = \{ w_1 ,..., w_j \} in the Markov network by summing over all possible assignments to u
otin V',W'; this is called Exact Inference . However, exact inference is in general a #P-complete problem, and thus computationally intractable. Approximation techniques such as Markov Chain Monte Carlo and loopy Belief Propagation are more feasible in practice. (Though note that some particular subclasses of MRF have polynomial algorithms; discovering such subclasses is an active research topic.)

One notable variant of a Markov network is a Conditional Random Field , in which each random variable may also be conditioned upon a set of global observations o. In this model, each function \phi_k is a mapping from all assignments to both the clique ''k'' and the observations o to the nonnegative real numbers. This form of the Markov network may be more appropriate for producing Discriminative Classifiers , which do not model the distribution over the observations.


SEE ALSO



EXTERNAL LINKS