GARIC architecture- Reinforcement Learning - FC

zwir Igor (ratto@datamar.com.ar)
Tue, 27 Aug 1996 14:57:20 +0200


Berenji show in his work a stochastic action in relation with
the internal reinforcement and the parameters learning method.
In a practical implementation I think there is an scaling factor
omited. If anybody know the relation, please answer me.