JointActionLearner.hpp

Go to the documentation of this file.

#ifndef AI_TOOLBOX_FACTORED_MDP_JOINT_ACTION_LEARNER_HEADER_FILE
#define AI_TOOLBOX_FACTORED_MDP_JOINT_ACTION_LEARNER_HEADER_FILE
 
#include <AIToolbox/MDP/Types.hpp>
#include <AIToolbox/Factored/MDP/Types.hpp>
 
#include <AIToolbox/Factored/Utils/Core.hpp>
 
#include <AIToolbox/MDP/Algorithms/QLearning.hpp>
 
namespace AIToolbox::Factored::MDP {
    class JointActionLearner {
        public:
            JointActionLearner(size_t S, Action A, size_t id, double discount = 1.0, double alpha = 0.1);
 
            void stepUpdateQ(size_t s, const Action & a, size_t s1, double rew);
 
            const AIToolbox::MDP::QFunction & getJointQFunction() const;
 
            const AIToolbox::MDP::QFunction & getSingleQFunction() const;
 
            void setLearningRate(double a);
 
            double getLearningRate() const;
 
            void setDiscount(double d);
 
            double getDiscount() const;
 
            size_t getS() const;
 
            const Action & getA() const;
 
            size_t getId() const;
 
        private:
            Action A;
            size_t id_;
 
            std::vector<unsigned> stateCounters_;
            boost::multi_array<std::vector<unsigned>, 2> stateActionCounts_;
 
            AIToolbox::MDP::QFunction singleQFun_;
            PartialFactorsEnumerator jointActions_;
 
            AIToolbox::MDP::QLearning qLearning_;
    };
}
 
#endif