It would have been a great lecture. Probably every one would have recommended to watch the video before diving into deep TD methods. The creator of TD himself explains the logic behind his algorithm, what a valuable content we lost because of the ignorant cameraman
Excellent lecture. This feels like a philosophy lesson and the equations just make sense! It is like a different view on the world, a more accurate view of the world; counter-intuitive at first but actually intuitive. He is just telling us his philosophy (model of the world), and it comes out as equations. Since he came up with the TD method, he can tell us exactly what he was thinking to come up with these equations. Such an honor to learn from him! [PS: All, use slides from description to follow along]
That wriggling dude is the one who introduced TD learning methods, and his text book is considered as the Bible for Reinforcement Learning :) But yes, poor camera work. They should have focused on the slides.
It would have been a great lecture. Probably every one would have recommended to watch the video before diving into deep TD methods. The creator of TD himself explains the logic behind his algorithm, what a valuable content we lost because of the ignorant cameraman
The legend himself ♥️
The legend himself.
his prediction about prediction learning at 8.38 turned out to be spot on.
Wish the camera focused in slides at least for the math
It's in the textbook pretty early on. Incompleteideas.net
There is a link to download the presentation (see Show More)
@@alexanderyau6347 The real mvp
Excellent lecture. This feels like a philosophy lesson and the equations just make sense! It is like a different view on the world, a more accurate view of the world; counter-intuitive at first but actually intuitive. He is just telling us his philosophy (model of the world), and it comes out as equations. Since he came up with the TD method, he can tell us exactly what he was thinking to come up with these equations. Such an honor to learn from him!
[PS: All, use slides from description to follow along]
It could have been great if you could provide the slides as well. since we don't see the board...
Oh no, he got stroke. So sad....hope you are doing well.
Great!
Adding slide number/header using timestamps would have been more helpful.
Camera is probably AI operated.
12:30
30:37
Maybe if i would have seen slides and not some dude wiggling, i would find it more interesting...
That wriggling dude is the one who introduced TD learning methods, and his text book is considered as the Bible for Reinforcement Learning :)
But yes, poor camera work. They should have focused on the slides.