Размер видео: 1280 X 720853 X 480640 X 360
Показать панель управления
Автовоспроизведение
Автоповтор
What a legendThanks so much you have a talent for teaching!
Thank you for this marvelous video! You explain these concepts so well!!!
@14:55 do we know R(s) or do we estimate it?
For ADP, we don't exactly know R. Though for this type of MDP, we can just memorize the R(s) we observe. In other MDPs, sometimes the reward can be randomized, so you can't just memorize it.
@14:25 The function seems incorrect according to my understanding. Since the policy is fixed. Why we need to select action.. I am very confused.
What a legend
Thanks so much you have a talent for teaching!
Thank you for this marvelous video! You explain these concepts so well!!!
@14:55 do we know R(s) or do we estimate it?
For ADP, we don't exactly know R. Though for this type of MDP, we can just memorize the R(s) we observe. In other MDPs, sometimes the reward can be randomized, so you can't just memorize it.
@14:25 The function seems incorrect according to my understanding. Since the policy is fixed. Why we need to select action.. I am very confused.