Deep RL Bootcamp Lecture 3: Deep Q-Networks

https://www.youtube.com/watch?v=fevMOp5TDQs

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

http://www.denizyuret.com/2015/03/alec-radfords-animations-for.html

 

 

 

 

 

artari is not a MDP, but MDP method works well. or use RNN

in many domains, people end up using RNN to represent q-function.

 

 

 

 

 

 

 replay really makes a difference!!!

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 should the two network have different set of hyperparameter? just like a group of workers with different kinds of personality? will the collaboration help?