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 CActorFromQFunctionAndPolicy Class Reference Actor which uses a QFunction and his Policy for the update. More... #include <cactorcritic.h> Inheritance diagram for CActorFromQFunctionAndPolicy: List of all members. Public Member Functions   CActorFromQFunctionAndPolicy (CAbstractQFunction *qFunction, CStochasticPolicy *policy)   Creates the actor object, the policy has to choose the actions using the specified Q-Function. virtual  ~CActorFromQFunctionAndPolicy () virtual void  receiveError (double critic, CStateCollection *state, CAction *Action, CActionData *data=NULL)   Updates the Q-Function. CStochasticPolicy *  getPolicy () Protected Attributes CStochasticPolicy *  policy double *  actionValues Detailed Description Actor which uses a QFunction and his Policy for the update. The only difference to CActorFromQFunction is the update of the Q-Function. The update is Q(s_t,a_t)_new = Q(s_t,a_t)_old + beta * td * (1 - pi_(s_t, a_t)), where pi(s_t, a_t) is the softmax-policy from the actor. This method is recommended by Sutton and Barto. Constructor & Destructor Documentation CActorFromQFunctionAndPolicy::CActorFromQFunctionAndPolicy (  CAbstractQFunction *  qFunction, CStochasticPolicy *  policy )    Creates the actor object, the policy has to choose the actions using the specified Q-Function. virtual CActorFromQFunctionAndPolicy::~CActorFromQFunctionAndPolicy (   )  [virtual] Member Function Documentation CStochasticPolicy* CActorFromQFunctionAndPolicy::getPolicy (   )  virtual void CActorFromQFunctionAndPolicy::receiveError (  double  critic, CStateCollection *  state, CAction *  Action, CActionData *  data = NULL )  [virtual]   Updates the Q-Function. Does the following update: Q(s_t,a_t)_new = Q(s_t,a_t)_old + beta * td * (1 - pi(s_t, a_t)) Reimplemented from CActorFromQFunction. Member Data Documentation double* CActorFromQFunctionAndPolicy::actionValues [protected] CStochasticPolicy* CActorFromQFunctionAndPolicy::policy [protected] The documentation for this class was generated from the following file: cactorcritic.h