But I think that in probabilistic considerations about bayes rule, the another assumption must be given, because if AB is an empty set, then P(A|B) is always equal to zero, so the integral int(P(A|B)dA also must be zero. Moreover - AB could be also a set of the zero measure - then we need another derivations.
It's really funny to me, that I never really learned completing the square, but rather just the formula to solve quadratic funktion and was told to learn it by heart which never stopped I guess (writing bachelor thesis in cs soon). Like how can our education system (germany) not teach me that lol. thats just mind blowing
![กี่หมื่นฟ้า | Your Sky Series EP.1 [2/4]](http://i.ytimg.com/vi/qJbQs_nhjD4/mqdefault.jpg)
Bayesian Learning starts at 20:51
Thank you for the course Professor Nando.
The corresponding homework also seems very educational. Can you please provide us with a link ?
But I think that in probabilistic considerations about bayes rule, the another assumption must be given, because if AB is an empty set, then P(A|B) is always equal to zero, so the integral int(P(A|B)dA also must be zero. Moreover - AB could be also a set of the zero measure - then we need another derivations.
What does P(A|B) mean if AB is an empty set?
Amazing example!
It's really funny to me, that I never really learned completing the square, but rather just the formula to solve quadratic funktion and was told to learn it by heart which never stopped I guess (writing bachelor thesis in cs soon). Like how can our education system (germany) not teach me that lol. thats just mind blowing
35:30 God tier explanation, ty
Does anyone know what was the book he recommended for this class? I couldn't hear the name of the author clearly.
www.inference.org.uk/itprnn/book.pdf
Machine Learning by Kevin Murphy. MIT Press. www.cs.ubc.ca/~murphyk/MLbook/index.html Mentioned at
about 8:53
Great man!
Start @15:00
At @35:00
thanks
Why are we not writing P(Y|XTHERE,SIGMA) instead of P(Y|X,THETA,SIGMA). X and theta would be multplited