MultiAgentDecisionProcess
|
QTableInterface is the abstract base class for Q(., a) functions. More...
#include <QTableInterface.h>
Public Member Functions | |
virtual QTableInterface * | Clone () const =0 |
Returns a pointer to a copy of this class. More... | |
virtual double | Get (Index s_i, Index ja_i) const =0 |
virtual void | Set (Index s_i, Index ja_i, double rew)=0 |
virtual | ~QTableInterface () |
QTableInterface is the abstract base class for Q(., a) functions.
It represents functions mapping from some domain (e.g. states, local states, histories, etc.) and some action domain (individual, joint or group actions) to a real number representing some form of payoff (long term reward, or immediate reward).
Note the argument of the functions defined here assume Q(s,a), but is should be clear that for s_i any general domain index may be used.
|
inlinevirtual |
|
pure virtual |
Returns a pointer to a copy of this class.
Implemented in RewardModelMappingSparseMapped, RewardModelMappingSparse, QTable, RewardModelMapping, and RewardModelDiscreteInterface.
Implemented in RewardModelMappingSparseMapped, RewardModelMappingSparse, RewardModelMapping, QTable, and RewardModelDiscreteInterface.
Implemented in RewardModelMappingSparseMapped, RewardModelMappingSparse, RewardModelMapping, QTable, and RewardModelDiscreteInterface.