Belief propagation
Algoritmo de propagación de creencias
- Date
- 1982 · confidence high
- Place
- Los Angeles, California, United States
- Credited to
- Judea Pearl
- Field
- Knowledge & symbols
What it is
Belief propagation is a message-passing algorithm for performing inference on probabilistic graphical models, such as Bayesian networks and Markov random fields, computing the marginal distribution of each unobserved variable. It was introduced by computer scientist Judea Pearl.
What it led to
- Basis for efficient inference in Bayesian networks
- Application in error-correcting codes (LDPC, turbo codes)
- Use in computer vision and natural language processing