Class LSTDOptions<T>

Namespace

AiDotNet.Models.Options

Assembly

AiDotNet.dll

Configuration options for LSTD (Least-Squares Temporal Difference) agents.

public class LSTDOptions<T> : ReinforcementLearningOptions<T>

Type Parameters

T

The numeric type used for calculations.

Inheritance

object

ReinforcementLearningOptions<T>

LSTDOptions<T>

Inherited Members

ReinforcementLearningOptions<T>.LearningRate

ReinforcementLearningOptions<T>.DiscountFactor

ReinforcementLearningOptions<T>.LossFunction

ReinforcementLearningOptions<T>.Seed

ReinforcementLearningOptions<T>.BatchSize

ReinforcementLearningOptions<T>.ReplayBufferSize

ReinforcementLearningOptions<T>.TargetUpdateFrequency

ReinforcementLearningOptions<T>.UsePrioritizedReplay

ReinforcementLearningOptions<T>.EpsilonStart

ReinforcementLearningOptions<T>.EpsilonEnd

ReinforcementLearningOptions<T>.EpsilonDecay

ReinforcementLearningOptions<T>.WarmupSteps

ReinforcementLearningOptions<T>.MaxGradientNorm

object.Equals(object)

object.Equals(object, object)

object.GetHashCode()

object.GetType()

object.MemberwiseClone()

object.ReferenceEquals(object, object)

object.ToString()

Remarks

LSTD solves for the optimal linear weights directly using matrix operations (A^-1 * b) rather than incremental updates. This provides more sample-efficient learning but requires solving a linear system.

For Beginners: LSTD is like solving a math equation directly instead of guessing and checking. It collects experiences and then computes the best weights all at once using linear algebra, rather than slowly adjusting them one step at a time.

Best for:

• Limited data scenarios (sample efficient)
• Batch learning from fixed datasets
• When you have computational power for matrix operations
• Problems where convergence speed matters

Not suitable for:

• Very large feature spaces (matrix becomes huge)
• Online learning (needs batches)
• When computational resources are limited
• Non-linear function approximation needs

Properties

ActionSize

Size of the action space (number of possible actions).

public int ActionSize { get; init; }

Property Value

int

FeatureSize

Number of features in the state representation.

public int FeatureSize { get; init; }

Property Value

int

RegularizationParam

Regularization parameter to prevent overfitting and ensure numerical stability.

public double RegularizationParam { get; init; }

Property Value

double