Next:
TD-Gammon
TD-Gammon
MDPs with a very large number of states
State encoding TD-Gammon
MDP description
"Encoding" Functions
methods of encoding
Choosing the parameters for
F
(
r
,
s
)
TD-Gammon
Calculating the Derivative of a Neural Network
About this document ...
Yishay Mansour
2000-01-17