RobustnessResult<T>

Models & Types · AiDotNet.Evaluation.Engines

Results from robustness analysis.

Constructors

Constructor Summary
RobustnessResult()

Properties

Property Summary
Dictionary<double, double> DropoutRobustness { get; set; } Performance at each dropout rate (dropout rate → score).
Dictionary<double, double> NoiseDegradation { get; set; } Performance degradation at each noise level.
Dictionary<double, double> NoiseRobustness { get; set; } Performance at each noise level (noise level → score).
Dictionary<int, double> FeatureImportance { get; set; } Feature importance via permutation (feature index → importance).
double BaselineScore { get; set; } Baseline score without perturbations.
double OverallRobustnessScore { get; set; } Overall robustness score (0 to 1, higher is better).
int NumFeatures { get; set; } Number of features in the dataset.
int NumSamples { get; set; } Number of samples tested.
string MetricName { get; set; } Name of the metric tracked.