On the Convergence of Loopy Belief Propagation Algorithm for Different Update Rules.
Nobuyuki TagaShigeru MasePublished in: IEICE Trans. Fundam. Electron. Commun. Comput. Sci. (2006)
Keyphrases
- loopy belief propagation
- dynamic programming
- update rules
- learning algorithm
- convergence rate
- k means
- update rule
- free energy
- expectation maximization
- worst case
- optimization algorithm
- conditional random fields
- negative matrix factorization
- prior information
- data clustering
- fixed point
- closed form
- least squares
- probabilistic model
- optimal solution
- objective function