AdversarialAttackOptions<T>

Options & Configuration · AiDotNet.Models.Options

Configuration options for adversarial attack algorithms.

Inheritance: ModelOptions → AdversarialAttackOptions<T>

Known derived types: AutoAttackOptions<T>

For Beginners

These settings control how the “stress test” for your AI works. You can adjust how strong the attacks are, how many attempts they make to fool your model, and what type of changes they’re allowed to make to inputs.

How It Works

These options control how adversarial examples are generated, including the strength of perturbations, attack iterations, and norm constraints.

Constructors

Constructor Summary
AdversarialAttackOptions()

Properties

Property Summary
AdversarialNormType NormType { get; set; } Gets or sets the norm type for perturbation constraints.
bool IsTargeted { get; set; } Gets or sets whether to use targeted or untargeted attacks.
bool UseRandomStart { get; set; } Gets or sets whether to use random initialization.
double CWConfidence { get; set; } Gets or sets the Carlini-Wagner confidence margin kappa.
double CWConstant { get; set; } Gets or sets the Carlini-Wagner trade-off constant c between distortion and the attack loss.
double CWLearningRate { get; set; } Gets or sets the Adam learning rate the Carlini-Wagner attack uses in tanh space.
double Epsilon { get; set; } Gets or sets the maximum perturbation budget (epsilon).
double StepSize { get; set; } Gets or sets the step size for iterative attacks.
int Iterations { get; set; } Gets or sets the number of iterations for iterative attacks.
int RandomSeed { get; set; } Gets or sets the random seed for reproducibility.
int TargetClass { get; set; } Gets or sets the target class for targeted attacks.