Belief Propagation

preview_player
Показать описание
Virginia Tech
Machine Learning

Two corrections:

1. At 5:48, it should be m_{s to t}(x_t), not m_{t to s}(x_s).
2. At 7:22, the potential in the example message should be \phi(X_B, X_D), not \phi(X_B, X_C).
Рекомендации по теме
Комментарии
Автор

There is a typo in 7:29, when calculating m_{B\to D}(X_D), it should be phi(X_B, X_D).

JirongYi
Автор

Another typo starting at 5:48? Should be m_s->t(x_t), not m_t->s(x_s)

gabrielorstadius
Автор

Excellent video! I hope that you do more of these videos for graphical models

vincenttan
Автор

In my class the professor said that loopy belief propagation is not guaranteed to converge.

AltairS
Автор

what does it mean "unordered vectors of random variable and trying to index into them" at 3:30

wahabfiles
Автор

Very good! I finally understood it, thanks man

darmstadtbeste
Автор

Is there a blue "down" arrow missing for the right most node at 18:10? Correct me if I'm wrong.

GaGa
Автор

Fantastic video
Thanks so much for making this

stevendurr
Автор

is this same as max product algorithm?

justinbieber