This was amazing! Struggling to understand the conundrum of MRFs. This video is the only good resource I've come across which explains it so clearly. Thanks. 😊
Hi! Nice video! What I can not understand is what is a "potential function" ? From what I see, they do not represent a distribution since they do not sum up to 1, but what I can not understand is how do we define these functions...Could you explain it?
A clique is a set of nodes that are all directly connected to each other. A maximal clique is a clique that is not contained in a larger clique. E.g., a triangle ABC is a clique, but so is AB. ABC contains AB, so AB can't be a maximal clique.
![Noob To Max With DRAGON REWORK In Blox Fruits [FULL MOVIE]](http://i.ytimg.com/vi/LnBMOinoOvA/mqdefault.jpg)
I've been taught MRFs twice now, and I have to say this is the most informative way I've had it taught. Great job!
on 6:50, why not {B} included in the separating subset?
It can be, too!
What a good channel i've come across. This will definitely become a part of my library to extend my problem solving toolbox as a developer.
Thank you!
This was amazing! Struggling to understand the conundrum of MRFs. This video is the only good resource I've come across which explains it so clearly. Thanks. 😊
A great explanation at 16:48! Bayes Net can be converted to MRF but lose marginal independence when a shared child occurs.
I need to read more about Markov Random Field in detail, can you suggest any resource?
Thank you Professor Huang.
11:49 Can anyone tell me why phi(A, C) = P(C|A) and not P(A)P(C|A)?
Thanks for helping me❤❤❤
Hi! Nice video! What I can not understand is what is a "potential function" ? From what I see, they do not represent a distribution since they do not sum up to 1, but what I can not understand is how do we define these functions...Could you explain it?
Hi, the Z (sum of potential of cliques) denominator will get the sum to 1
Thanks for great explanation. One question: why bother converting a BN into an MRF?
You would be able to use a general algorithm of message passing for inference that holds for junction trees
Great Video!
Really good lecture!!
Simple and clear!
what is 'phi' sir
What is a maximal clique?
A clique is a set of nodes that are all directly connected to each other. A maximal clique is a clique that is not contained in a larger clique. E.g., a triangle ABC is a clique, but so is AB. ABC contains AB, so AB can't be a maximal clique.
@@berty38
Makes so much sense! Awesome explanation.