LogCoshLossMetric<T>
Models & Types · AiDotNet.Evaluation.Metrics.Regression
Computes Log-Cosh Loss: mean of log(cosh(y - ŷ)).
Implements: IRegressionMetric<T>
For Beginners
Log-Cosh loss combines benefits of MSE and MAE:
- Behaves like MAE for large errors (robust to outliers)
- Behaves like MSE for small errors (smooth gradient)
- Always positive and convex
- Has continuous second derivative (unlike Huber)
How It Works
Log-Cosh = (1/N) × Σ log(cosh(y - ŷ))
Advantages over alternatives:
- vs MSE: More robust to outliers
- vs MAE: Smoother gradient near zero
- vs Huber: No hyperparameter to tune
Constructors
| Constructor | Summary |
|---|---|
LogCoshLossMetric() |
Properties
| Property | Summary |
|---|---|
MetricDirection Direction { get; } |
|
T MaxValue { get; } |
|
T MinValue { get; } |
|
bool RequiresProbabilities { get; } |
|
bool SupportsMultiClass { get; } |
|
string Category { get; } |
|
string Description { get; } |
|
string Name { get; } |
Methods
| Method | Summary |
|---|---|
MetricWithCI<T> ComputeWithCI(ReadOnlySpan<T> predictions, ReadOnlySpan<T> actuals, ConfidenceIntervalMethod ciMethod, double confidenceLevel, int bootstrapSamples, int? randomSeed) |
|
T Compute(ReadOnlySpan<T> predictions, ReadOnlySpan<T> actuals) |