GridworldDRLExperienceReplayPOMDP.hpp

Go to the documentation of this file.

#ifndef SRC_APPLICATION_GRIDWORLDDRLEXPERIENCEREPLAYPOMDP_HPP_
#define SRC_APPLICATION_GRIDWORLDDRLEXPERIENCEREPLAYPOMDP_HPP_

#include <vector>
#include <string>

#include <opengl/application/OpenGLEpisodicApplication.hpp>
#include <opengl/visualization/WindowCollectorChart.hpp>
using namespace mic::opengl::visualization;

#include <mlnn/BackpropagationNeuralNetwork.hpp>
// Using multi layer neural networks
using namespace mic::mlnn;
using namespace mic::types;

#include <types/Gridworld.hpp>
#include <types/SpatialExperienceMemory.hpp>

namespace mic {
namespace application {


class GridworldDRLExperienceReplayPOMDP: public mic::opengl::application::OpenGLEpisodicApplication {
public:
    GridworldDRLExperienceReplayPOMDP(std::string node_name_ = "application");

    virtual ~GridworldDRLExperienceReplayPOMDP();

protected:

    virtual void initialize(int argc, char* argv[]);

    virtual void initializePropertyDependentVariables();

    virtual bool performSingleStep();

    virtual void startNewEpisode();

    virtual void finishCurrentEpisode();


private:

    WindowCollectorChart<float>* w_chart;

    mic::utils::DataCollectorPtr<std::string, float> collector_ptr;

    mic::environments::Gridworld grid_env;

    size_t batch_size;

    mic::configuration::Property<float> step_reward;

    mic::configuration::Property<float> discount_rate;

    mic::configuration::Property<float> learning_rate;

    mic::configuration::Property<double> epsilon;

    mic::configuration::Property<int> step_limit;

    mic::configuration::Property<std::string> statistics_filename;

    mic::configuration::Property<std::string> mlnn_filename;

    mic::configuration::Property<bool> mlnn_save;

    mic::configuration::Property<bool> mlnn_load;

    BackpropagationNeuralNetwork<float> neural_net;

    float computeBestValueForGivenStateAndPredictions(mic::types::Position2D player_position_, float* predictions_);

    mic::types::MatrixXfPtr getPredictedRewardsForGivenState(mic::types::Position2D player_position_);

    mic::types::NESWAction selectBestActionForGivenState(mic::types::Position2D player_position_);

    std::string streamNetworkResponseTable();

    long long sum_of_iterations;

    long long sum_of_rewards;

    long long number_of_successes;

    SpatialExperienceMemory experiences;
};

} /* namespace application */
} /* namespace mic */

#endif /* SRC_APPLICATION_GRIDWORLDDRLEXPERIENCEREPLAYPOMDP_HPP_ */