CertifiedDefenseOptions<T>
Options & Configuration · AiDotNet.Models.Options
Configuration options for certified defense mechanisms.
Inheritance: ModelOptions → CertifiedDefenseOptions<T>
For Beginners
These settings control how the “guaranteed protection” works. You can adjust how many samples to use, how tight the guarantees should be, and what certification method to apply.
How It Works
These options control certified robustness methods that provide provable guarantees about model predictions under adversarial perturbations.
Constructors
| Constructor | Summary |
|---|---|
CertifiedDefenseOptions() |
Properties
| Property | Summary |
|---|---|
AdversarialNormType NormType { get; set; } |
Gets or sets the norm type for certification. |
CertifiedDefenseMethod CertificationMethod { get; set; } |
Gets or sets the certification method to use. |
bool UseTightBounds { get; set; } |
Gets or sets whether to use tight bounds computation. |
double ConfidenceLevel { get; set; } |
Gets or sets the confidence level for certification. |
double NoiseSigma { get; set; } |
Gets or sets the noise standard deviation for randomized smoothing. |
int BatchSize { get; set; } |
Gets or sets the batch size for certification. |
int NumSamples { get; set; } |
Gets or sets the number of samples for randomized smoothing. |
int SelectionSamples { get; set; } |
Gets or sets the number of noisy samples randomized smoothing uses to choose the top class before it estimates that class’s probability from NumSamples fresh samples. |