Class ReinforcementLearningOptions<T>

Namespace

AiDotNet.ReinforcementLearning.Agents

Assembly

AiDotNet.dll

Configuration options for reinforcement learning agents.

public class ReinforcementLearningOptions<T>

Type Parameters

T

The numeric type used for calculations.

Inheritance

object

ReinforcementLearningOptions<T>

Derived

A3COptions<T>

DecisionTransformerOptions<T>

DoubleQLearningOptions<T>

DreamerOptions<T>

DynaQOptions<T>

DynaQPlusOptions<T>

EpsilonGreedyBanditOptions<T>

ExpectedSARSAOptions<T>

GradientBanditOptions<T>

LSPIOptions<T>

LSTDOptions<T>

LinearQLearningOptions<T>

LinearSARSAOptions<T>

MADDPGOptions<T>

ModifiedPolicyIterationOptions<T>

MonteCarloExploringStartsOptions<T>

MonteCarloOptions<T>

MuZeroOptions<T>

NStepQLearningOptions<T>

NStepSARSAOptions<T>

OffPolicyMonteCarloOptions<T>

OnPolicyMonteCarloOptions<T>

PolicyIterationOptions<T>

PrioritizedSweepingOptions<T>

QLambdaOptions<T>

QMIXOptions<T>

RainbowDQNOptions<T>

SARSALambdaOptions<T>

SARSAOptions<T>

TD3Options<T>

TRPOOptions<T>

TabularActorCriticOptions<T>

TabularQLearningOptions<T>

ThompsonSamplingOptions<T>

UCBBanditOptions<T>

ValueIterationOptions<T>

WatkinsQLambdaOptions<T>

WorldModelsOptions<T>

Inherited Members

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

object.ToString()

Properties

BatchSize

Batch size for training updates.

public int BatchSize { get; init; }

Property Value

int

DiscountFactor

Discount factor (gamma) for future rewards.

public T? DiscountFactor { get; init; }

Property Value

T

EpsilonDecay

Exploration decay rate.

public double EpsilonDecay { get; init; }

Property Value

double

EpsilonEnd

Final exploration rate.

public double EpsilonEnd { get; init; }

Property Value

double

EpsilonStart

Initial exploration rate (for epsilon-greedy policies).

public double EpsilonStart { get; init; }

Property Value

double

LearningRate

Learning rate for gradient updates.

public T? LearningRate { get; init; }

Property Value

T

LossFunction

Loss function to use for training.

public ILossFunction<T>? LossFunction { get; init; }

Property Value

ILossFunction<T>

MaxGradientNorm

Maximum gradient norm for clipping (0 = no clipping).

public double MaxGradientNorm { get; init; }

Property Value

double

ReplayBufferSize

Size of the replay buffer (if applicable).

public int ReplayBufferSize { get; init; }

Property Value

int

Seed

Random seed for reproducibility (optional).

public int? Seed { get; init; }

Property Value

int?

TargetUpdateFrequency

Frequency of target network updates (if applicable).

public int TargetUpdateFrequency { get; init; }

Property Value

int

UsePrioritizedReplay

Whether to use prioritized experience replay.

public bool UsePrioritizedReplay { get; init; }

Property Value

bool

WarmupSteps

Number of warmup steps before training.

public int WarmupSteps { get; init; }

Property Value

int