IMetric<T>

Interfaces · AiDotNet.Evaluation.Metrics

Base interface for all evaluation metrics.

Known implementations: IClassificationMetric<T>, IProbabilisticClassificationMetric<T>, IRankingMetric<T>, IRegressionMetric<T>, ITimeSeriesMetric<T>

Properties

Property Summary
MetricDirection Direction { get; } Gets whether higher values indicate better performance.
T MaxValue { get; } Gets the maximum possible value for this metric, if bounded.
T MinValue { get; } Gets the minimum possible value for this metric, if bounded.
bool RequiresProbabilities { get; } Gets whether this metric requires probability predictions (not just class labels).
bool SupportsMultiClass { get; } Gets whether this metric is suitable for multi-class classification.
string Category { get; } Gets the category of the metric (Classification, Regression, etc.).
string Description { get; } Gets a human-readable description of the metric.
string Name { get; } Gets the unique name of the metric.