Brilliant! Examples are great, easy to follow and understand! One thing I want to point out is that the V structure has opposite behavior compare to other paths.
Correct, for the conditional independence to hold, ALL paths must be inactive. So the algorithm requires to keep checking paths until either an active path has been found --- in which case we cannot claim the conditional independence; or until all paths have been checked and all have been found inactive --- in which case we can claim the conditional independence.
Sure. It's actually much simpler. It is guaranteed that X is independent of Y given {Z_1, ..., Z_m} if every path that connects X and Y has at least one of the Z_i variables in it.
even after 12 years, its useful, thank you
eight years later this is still useful, thx
did u use this 8 years after college?
@@lazyprodigy yes
@@jorgav1794 Can i have some context? i am curious now :))
@@lazyprodigy well, I took a Data Science course and this was one of the topics :P
Brilliant! Examples are great, easy to follow and understand! One thing I want to point out is that the V structure has opposite behavior compare to other paths.
you explained it sooo much better than my tutor. thanks a lot!
Thank you for the explanation, it was much easier to understand this from you!
Thank you so much, Sir. God bless you.
You made it very easy for me to understand.. thank you!
Thank u very much just a few min ago I did'nt know any thing about d-seperation but now I can easily handel the related questions.
@Aram IR -- Happy to read it was helpful!
Correct, for the conditional independence to hold, ALL paths must be inactive. So the algorithm requires to keep checking paths until either an active path has been found --- in which case we cannot claim the conditional independence; or until all paths have been checked and all have been found inactive --- in which case we can claim the conditional independence.
Thank You so much for the great explanation.
So good! So much better than my ******* professor
THANK YOU SO SO MACH
Would you please explain how can we find independencies in markov networks? Can we use the same technique?
Sure. It's actually much simpler. It is guaranteed that X is independent of Y given {Z_1, ..., Z_m} if every path that connects X and Y has at least one of the Z_i variables in it.
From which course is this slide from?
this is the only explanation on youtube i understand. the rest were useless