Reinforcement Learning Toolbox 2.0 last updated:

General Documentation Manual Tutorial Class Reference Master Thesis Examples Related Papers Downloads Links News mailto:webmaster Main Page     Class Hierarchy   Compound List   File List   Compound Members   File Members CActor Class Reference Interface for all Actors. More... #include <cactorcritic.h> Inheritance diagram for CActor: List of all members. Public Member Functions   CActor () virtual void  receiveError (double critic, CStateCollection *oldState, CAction *Action, CActionData *data=NULL)=0   interface function for the actors double  getLearningRate () void  setLearningRate (double learningRate) Detailed Description Interface for all Actors. The actors have to adopt their policies according to the critics they get. The class CActor provides an interface for sending a critic to the actor (receiveError(CStateCollection *currentState, CAction *action). This is all the Actor classes have to implement. How to get the policy from the actor is decontrolled to the Actor classes itself. The class also maintains a learning rate beta, which can be used by the subclasses. Constructor & Destructor Documentation CActor::CActor (   )  Member Function Documentation double CActor::getLearningRate (   )  virtual void CActor::receiveError (  double  critic, CStateCollection *  oldState, CAction *  Action, CActionData *  data = NULL )  [pure virtual]   interface function for the actors The actor gets a critic for a given state action pair. Then he has to adopt his policy according to the critic. Implements CErrorListener. Implemented in CActorFromQFunction, CActorFromQFunctionAndPolicy, CActorFromActionValue, CActorFromContinuousActionGradientPolicy, and CActorForMultipleAgents. void CActor::setLearningRate (  double  learningRate  )  The documentation for this class was generated from the following file: cactorcritic.h