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 RLToolbox Class List Here are the classes, structs, unions and interfaces with brief descriptions: C1DSplittingCondition CAbsoluteSoftMaxDistribution CAbstractBetaCalculator CAbstractFeatureStochasticEstimatedModel Base class for all estimated models CAbstractFeatureStochasticModel Interface for all model classes CAbstractQETraces Interface for Q-ETraces CAbstractQFunction Interface for all Q-Functions CAbstractStateDiscretizer Interface for all state discretizer CAbstractVETraces Class representing etraces for a V-Function CAbstractVFunction Interface reprenting a Value Function CAcroBotExpRewardFunction CAcroBotHeightRewardFunction CAcroBotModel CAcroBotRewardFunction CAcroBotVelocityRewardFunction CAction Class Representing an action the agent or any other SMDP can choose from CActionData Interface for saving changable data of an action CActionDataSet Class for mantaining the action data objects of all actions of an action set CActionDistribution Action Distribution classes define the distributions of stochastic Policies CActionList Class for logging a sequence of actions CActionObject Base Class for all Classes which have to maintain an action set CActionOutput CActionSet Class maintaining all the actions available for a certain object (normally a CActionObject like a controller) CActionStatistics Actionstatistics - for comparison of policies and agent controllers CActionStatisticsComparator Baseclass for all ActionStatisticComparators, used by CActor Interface for all Actors CActorForMultipleAgents CActorFromActionValue Actor class which can only decide beween 2 different action, depending on the action value of the current state CActorFromContinuousActionGradientPolicy CActorFromQFunction Actor who creates his Policy on a Q Function CActorFromQFunctionAndPolicy Actor which uses a QFunction and his Policy for the update CAdaptiveEtaCalculator Adaptive Learning Rate Calculator Interface class CAdaptiveParameterBoundedValuesCalculator Super class for all classes which use bounded target values CAdaptiveParameterCalculator Interface for all adaptive Parameter Calculators CAdaptiveParameterFromAverageRewardCalculator Adaptive Parameter Calculator which calculates the parameter's value from the current average reward CAdaptiveParameterFromNEpisodesCalculator Adaptive Parameter Calculator which calculates the parameter's value from the number of learning episodes CAdaptiveParameterFromNStepsCalculator Adaptive Parameter Calculator which calculates the parameter's value from the number of learning steps CAdaptiveParameterFromValueCalculator Adaptive Parameter Calculator which calculates the parameter's value from the current value of a V-Function CAdaptiveParameterUnBoundedValuesCalculator Super class for all classes which use unbounded target values CAdaptiveTargetGraph CAdaptiveTargetGraphController CAdvantageLearner CAdvantageUpdating CAgent The class represents the main acting Object of the Learning System, the agent CAgentController Class for navigating the agent CAgentLogger Class for logging a whole training trial CAgentStatisticController Agent controller which returns additionally an statistic object for the action CAverageReinforcementBaseLineCalculator CAverageRewardCalculator CAverageRewardSameStateCalculator CAverageRewardTestSuiteEvaluator CBatchCAQDataGenerator CBatchDataGenerator CBatchEpisodeUpdate Class for doing batch updates during the learning trial CBatchGradientLearner CBatchLearningPolicy CBatchQDataGenerator CBatchStepUpdate Class for doing Batch Updates CBatchVDataGenerator CCAGradientPolicyInputDerivationCalculator CCAGradientPolicyNumericInputDerivationCalculator CCALinearFAQETraces CCALinearFAQFunction CCartPoleHeightRewardFunction CCartPoleModel CCartPoleRewardFunction CComposedQETraces This is the E-Trace class for the CComposedQFunction class CComposedQFunction CComposedTransitionFunction CConjugateGradientLearner CConstantBetaCalculator CConstantGradientFunctionUpdater CConstantReinforcementBaseLineCalculator CContinuousAction CContinuousActionAddController CContinuousActionController CContinuousActionData Class for saving the continuous Values from a ContinuosAction CContinuousActionFeaturePolicy CContinuousActionGradientPolicy CContinuousActionLinearFA CContinuousActionPolicy CContinuousActionPolicyFromGradientFunction CContinuousActionProperties CContinuousActionQFunction CContinuousActionRandomPolicy CContinuousActionSigmoidPolicy CContinuousActionSmoother CContinuousCoulomResidual CContinuousDynamicProgramming CContinuousDynamicQProgramming CContinuousDynamicVProgramming CContinuousEulerResidual CContinuousMCQEvaluation CContinuousRBFAction CContinuousStateList CContinuousStateRegion CContinuousTimeAndActionBangBangVMPolicy CContinuousTimeAndActionSigmoidVMGradientPolicy CContinuousTimeAndActionSigmoidVMPolicy CContinuousTimeAndActionTransitionFunction CContinuousTimeAndActionVMPolicy CContinuousTimeParameters CContinuousTimeQFunctionFromTransitionFunction CContinuousTimeTransitionFunction CContinuousTimeVMPolicy CControllerAnalyzer Controller analyzer CDataCollector CDataCollectorFromAgentLogger CDataPreprocessor CDataSet CDataSet1D CDeterministicController Controller class makes a given controller deterministic CDirectGradient CDiscreteResidual CDiscreteStateOperatorAnd The and operator combines several discrete states to one discrete state CDiscreteStochasticEstimatedModel Estimated Model for Discrete States CDistributions Collection of Distributions CDivergentQFunctionException This exception is thrown if a value function has become divergent CDivergentVFunctionException This exception is thrown if a value function has become divergent CDynamicLinearContinuousTimeModel CDynamicProgramming Collection of static functions for dynamic Programming CEnvironmentModel The environment in which the agent can act CEpisode Class for logging a single Episode CEpisodeHistory Interface for all classes which are able to store episodes CEpisodeHistorySubset CEpisodeMatlabOutput CEpisodeOutput This Class writes each step and start of a new episode in readable form to a file CEpisodeOutputStateChanged Class writes only the steps in which the specified state changes in readable form to a file CEpsilonGreedyDistribution Class for the epsilon greedy action distribution CErrorListener CErrorSender CEvaluator CExplorationQFunction CExtendedAction This abstract class represents extended actions like behaviors or hierarchical SMDPs CExtendedActionTransitionFunction CExtendedPrimitiveAction CExtraLinearRegressionModelForestLearner CExtraLinearRegressionModelTree CExtraModelTree CExtraRegressionForestFeatureLearner CExtraRegressionForestLearner CExtraRegressionForestTrainer CExtraRegressionTree CExtraTree< TreeData > CExtraTreeRBFLinearWeightForest CExtraTreeRegressionForest CExtraTreesSplittingConditionFactory CFeature Class for storing a single feature (feature Index and feature Faktor CFeatureCalculator Base class for all Feature Calculators CFeatureFunction The feature function for storing features in an double array CFeatureList Class for storing features as sparse array CFeatureMap CFeatureOperatorAnd Combines 2 or more feature states with an "And" operatation CFeatureOperatorOr Combines 2 or more feature states with an "Or" operatation CFeatureQFunction Composed feature Q-Function CFeatureRewardFunction Class for calculating the reward given a feature not a state CFeatureRewardFunctionFromValueFunction CFeatureRewardModel CFeatureStateNNInput CFeatureStateRewardFunction CFeatureStateRewardModel CFeatureStochasticEstimatedModel Estimated Model for feature States CFeatureStochasticModel Class for loading and storing a fixed Model CFeatureVETraces This class is used as ETraces for feature V-Functions CFeatureVFunction CFeatureVRegressionTreeFunction CFittedCAQIteration CFittedIteration CFittedQIteration CFittedQIterationAnalyzer CFittedQNewFeatureCalculator CFittedVIteration CForest< TreeData > ***************** CForest *********************** CForestFeatureCalculator< TreeData > CFunctionComperator Super class of the V-Function Comperators and Q-Function comperators CGlobalGridWorldDiscreteState CGPOMDPGradientCalculator CGradientCalculator CGradientFunction Interface for all functions which support gradient update and gradient and output calculation CGradientFunctionUpdater CGradientLearner CGradientQETraces E-Trace class for all gradient Q-Functions CGradientQFunction CGradientUpdateFunction Interface class for all function which gradient update CGradientVETraces ETraces for gradient functions CGradientVFunction Interface for all classes that can use gradients for updating CGraphAdaptiveTargetDynamicProgramming CGraphController CGraphDebugger CGraphDynamicProgramming CGraphLogger CGraphTarget CGraphTransition CGraphTransitionAdaptiveTarget CGraphValueFromValueFunctionCalculator CGreedyASComparator Comparator for CMulticontrollerGreedyPolicy prefers best action CGreedyDistribution Class for a greedy action distribution CGridFeatureCalculator Abstract Superclass for all feature calculator that partition the state space with a grid CGridWorld CGridWorldAction CGridWorldController CGridWorldController::GridControllerRecord CGridWorldModel CGSearchPolicyGradientUpdater CHiearchicalAgent CHierarchicalController Class for calculating a hierarchical execution of a hierarchical structure CHierarchicalSemiMarkovDecisionProcess Subclass of CSemiMarkovDecisionProcess, used for hierarchical Learning CHierarchicalStack A list of Actions representing the actual hierarchical Stack CHierarchicalStackEpisode Class for logging the Hierarchical Stack of a training trial CHierarchicalStackListener Listener who gets the Hierarchical Stack instead of an action CHierarchicalStackSender Base class for sending the hierarchical stack to the listeners CIndividualEtaCalculator Eta Calculator, which allows you to set individual learning rates for each weight CKDRectangle CKDTree CKDTreeMedianSplittingFactory CKNearestLeaves< TreeData > CKNearestNeighbors CKNearestNeighborsTreeData< DataElement, TreeData > CKNearestRBFCenters CLeaf< TreeData > CLeafIndexFactory CLearnDataObject Interface for all objects that learn, like V-Functions or Q-Functions CLeastSquaresLearner CLinearActionContinuousTimeTransitionFunction CLinearFAContinuousAction CLinearInterpolationFeatureCalculator CLinearMultiFeatureCalculator Super class for all feature calculators that use a grid and where more than one feature can be active CLinearRegression CLinearRegressionDataFactory CLineSearchGradientFunctionUpdater CListenerTestSuite CLocal4GridWorldState CLocal4XGridWorldState CLocal8GridWorldState CLocalGridWorldDiscreteState CLocalLinearLearner CLocalLinearRegression CLocalRBFLearner CLocalRBFRegression CLocalRegression CLSTDLambda CLSTDOfflineEpisodePolicyEvaluation CLSTDOnlinePolicyEvaluation CMapping< OutputValue > CMatlabEpisodeOutputLogger CMatlabQAnalyzerLogger CMatlabVAnalyzerLogger CMeanStdPreprocessor CModelStateDiscretizer Class calculating a single discrete state from several state variables CModelTree CMonteCarloCAQLearner CMonteCarloError CMonteCarloQError CMonteCarloQLearner CMonteCarloSupervisedLearner CMonteCarloVError CMonteCarloVLearner CMultiPoleAction CMultiPoleContinuousReward CMultiPoleController CMultiPoleDiscreteController CMultiPoleDiscreteState CMultiPoleFailedState CMultiPoleModel CMultiStepAction CMultiStepActionData Class for saving the duration and the finished flag from a MultiStepAction CMyArray< T1 > A multidimensional Array CMyArray2D< T1 > 2 dimensional Array CMyArray3D< T1 > 3 dimensional array CMyException Interface for all Exception classes used by the toolbox CNeuralNetworkStateModifier State Modifier used for Neural Networks input states CNewFeatureCalculator CNewFeatureCalculatorDataGenerator CNode< TreeData > CNumericPolicyGradientCalculator COfflineEpisodePolicyEvaluation COnlinePolicyEvaluation COptimalVFunctionFromQFunction CParameterObject Super class of all Objects mantaining parameters CParameters This class represents a parameterset CPEGASUSAnalyticalPolicyGradientCalculator CPEGASUSNumericPolicyGradientCalculator CPEGASUSPolicyGradientCalculator CPEGreedyASComparator Comparator for CMulticontrollerGreedyPolicy prefers epsilon-greedy action over best action CPendulumModel CPendulumRewardFunction CPendulumUpTimeCalculator CPickupAction CPOASComparator Comparator for CMulticontrollerGreedyPolicy prefers actions of a specific owner CPolicyEvaluation CPolicyEvaluationGradientFunction CPolicyEvaluationTestSuite CPolicyEvaluator CPolicyGradientCalculator CPolicyGradientTestSuite CPolicyGradientWeightDecayListener CPolicyGreedynessEvaluator CPolicyIteration CPolicyIterationNewFeatures CPolicyIterationTestSuite CPolicySameStateEvaluator CPrimitiveAction CPrimitiveActionStateChange This class represents an primitive action which gets executed until a specific (mostly discrete) state changes CPrioritizedSweeping Class for model based Prioritized Sweeping CPutdownAction CQAverageTDErrorLearner CQAverageTDVarianceLearner CQETraces Q-ETraces consisting of V-Etraces (one for each action) CQFunction Compounded Q-Function consisting of V-Funcions CQFunctionAnalyzer Analyzer for Q-Functions CQFunctionComperator Comperator for 2 Q-Functions CQFunctionFromGradientFunction CQFunctionFromStochasticModel Converts a VFunction and a Model to a Q-Function CQFunctionFromTransitionFunction CQFunctionSum CQGreedyPolicy Greedy Policy based on a Q-Function CQLearner Class for Q-Learning CQLSTDLambda CQStochasticExplorationPolicy CQStochasticPolicy Stochastic Policy which computes its propabilities from the Q-Values of CRaceTrack CRaceTrackDiscreteState CRandomMaxPolicyGradientCalculator CRandomPolicyGradientCalculator CRangeSearch CRBFBasisFunction CRBFBasisFunctionLinearWeight CRBFCenterFeatureCalculator CRBFCenterNetwork CRBFCenterNetworkKDTree CRBFCenterNetworkSimpleSearch CRBFDataFactory CRBFExtraRegressionForest CRBFExtraRegressionTree CRBFFeatureCalculator Represents a normalized RBF network CRBFForestLearner CRBFLinearWeightDataFactory CRBFLinearWeightForest CRBFRegressionTreeOutputMapping CRegion CRegressionFactory CRegressionForest ***************** CRegressionForest *********************** CRegressionMultiMapping CRegressionTreeFunction CRegressionTreeQFunction CRegressionTreeVFunction CREINFORCELearner CReinforcementBaseLineCalculator CResidualBetaFunction CResidualFunction CResidualGradientFunction CRewardAsVFunction CRewardEpisode Logs the reward during an Episode CRewardFunction Class for clalculating the reward for the learning objects CRewardFunctionFromValueFunction CRewardHistory CRewardHistorySubset CRewardLogger CRewardPerEpisodeCalculator CSampleTransition CSamplingBasedGraph CSamplingBasedTransitionModel CSamplingBasedTransitionModelFromTransitionFunction CSarsaLearner Class for Sarsa Learning CSelectiveExplorationCalculator CSemiMarkovDecisionProcess Class for providing the general Functions for the learning Environment CSemiMDPLastNRewardFunction Reward Function for Behaviours CSemiMDPListener Interface for all SemiMDP Listeners CSemiMDPRewardListener Represents SMDP Listener which also need a reward CSemiMDPSender Class for sending the State-Action-State Tuple to the Listeners CSemiMDPTransition Transition for the semiMDP case CSingleStateDiscretizer Discretizes the original States i'th continuous state into the specified partitions CSingleStateFeatureCalculator Superclass for calculating features from a single continuous state variable CSingleStateLinearInterpolationFeatureCalculator Class for linear interpolation features of a single continuous state variable CSingleStateRBFFeatureCalculator Represents a 1-D RBF-network layed over a specified continous state variable CSmallLocalGridWorldDiscreteState CSoftMaxDistribution Soft Max Distribution for Stochastic Policies CSplittingCondition CSplittingConditionFactory CState Represents a single state CStateActionErrorLearner CStateActionTransitions Class for storing the Backward and Forward Transitions for a given state action pair CStateCollection Interface for a state collection CStateCollectionImpl Implementation of the CStateCollection interface CStateCollectionList Class for storing a sequence of state collections CStateErrorLearner CStateList Class for logging a list of States CStateModifier Abstract class for calculating modified states from the original Model state CStateModifiersObject Base Class for all Classes that have to maintain a list of modifiers CStateMultiModifier Interface for all state modifier how have acces to several other state modifier for state calculation CStateObject Base class for all classes which have to manage with states CStateOutput CStateProperties Class defining the Properties of a State CStateReward Reward Function that only depends on the current state CStateVariablesChooser CStateVETraces State ETraces stores the state object itself CStaticContinuousAction CStep Class storing a single step (State-Action-State Tuple) CStepHistory Class mantaining an unordered set of steps, which can be used for Learning CStochasticModelAction CStochasticPolicy Class for modeling a stochastic policy CStoredEpisodeModel Serves as environment model for an agent, simulating the episodes from an agent logger CSubGoalBehaviour CSubGoalController CSubGoalOutput CSubsetFactory CSupervisedFeatureGradientCalculator CSupervisedGradientCalculator CSupervisedGradientLearner CSupervisedLearner CSupervisedNeuralNetworkMatlabLearner CSupervisedNeuralNetworkTorchLearner CSupervisedQFunctionLearner CSupervisedQFunctionLearnerFromLearners CSupervisedQFunctionWeightedLearner CSupervisedQFunctionWeightedLearnerFromLearners CSupervisedWeightedLearner CTaxiDomain CTaxiHierarchicalBehaviour CTaxiIsTargetDiscreteState CTDGradientLearner CTDLearner Class for temporal Difference Learning CTDResidualLearner CTestSuite CTestSuiteCollection CTestSuiteEvaluator CTestSuiteEvaluatorLogger CTestSuiteLoggerFromEvaluator CTilingFeatureCalculator Tilings represent a grid layed over the state space CTorchFunction Interface to integrate Torch Machines in the learning systems CTorchGradientEtaCalculator CTorchGradientFunction Class for learning with Torch-Gradient machines CTorchVFunction CTransition Transition for the Markov Case CTransitionFunction CTransitionFunctionEnvironment CTransitionFunctionFromStochasticModel CTransitionFunctionInputDerivationCalculator CTransitionFunctionNumericalInputDerivationCalculator CTransitionList Class for storing Transitions CTree< TreeData > CTreeDataFactory< TreeData > CTreeElement< TreeData > CUnknownDataQFunction CUnknownDataQFunctionFromLocalRBFRegression CValueCalculator CValueIteration The Value Iteration Algorithm CValueSameStateCalculator CVariableBetaCalculator CVarioEta Vario Eta Learning Rate Calculator CVAverageTDErrorLearner CVAverageTDVarianceLearner CVectorQuantizationFactory CVFunctionAnalyzer Analyzer for V-Functions CVFunctionComperator Comperator for 2 V-Functions CVFunctionFromGradientFunction Class for learning with Torch-Gradient machines CVFunctionFromQFunction CVFunctionGradientLearner CVFunctionInputDerivationCalculator Interface class for calculating the gradient dV(x)/dx CVFunctionLearner TD Learner for Value Function learning CVFunctionNumericInputDerivationCalculator Calculating the input derivation of a V-Function numerically CVFunctionResidualLearner CVFunctionSum CVisitStateActionCounter CVisitStateActionEstimator CVisitStateCounter CVLSTDLambda CVMStochasticPolicy Stochastic Policy which calculates its action from a Dynamic Model and a V-Function CVPolicyLearner CVTable Value Function as a table CZeroReward CZeroVFunction Value Function always returning zero DataSubset EvaluationValue