Class PrioritizedReplayConfig<T>

Namespace

AiDotNet.Configuration

Assembly

AiDotNet.dll

Configuration for prioritized experience replay (PER).

public class PrioritizedReplayConfig<T>

Type Parameters

T

The numeric type.

Inheritance

object

PrioritizedReplayConfig<T>

Inherited Members

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

object.ToString()

Remarks

For Beginners: PER samples experiences based on their TD-error (surprise). High-error experiences are sampled more often because they have more to teach.

Constructors

PrioritizedReplayConfig()

Creates a new instance with default values.

public PrioritizedReplayConfig()

Properties

Alpha

Alpha parameter controlling prioritization strength (0 = uniform, 1 = full prioritization).

public T Alpha { get; set; }

Property Value

T

Remarks

Common value: 0.6. Higher = stronger prioritization.

BetaAnnealingSteps

Steps over which to anneal beta from initial to final.

public int BetaAnnealingSteps { get; set; }

Property Value

int

FinalBeta

Final beta value (should reach 1.0 by end of training).

public T FinalBeta { get; set; }

Property Value

T

InitialBeta

Initial beta for importance sampling correction.

public T InitialBeta { get; set; }

Property Value

T

Remarks

Starts low and anneals to 1.0. Common initial: 0.4.

PriorityEpsilon

Small constant added to priorities to prevent zero probabilities.

public T PriorityEpsilon { get; set; }

Property Value

T