ValidationCurveOptions
Options & Configuration · AiDotNet.Evaluation.Options
Configuration options for validation curve analysis.
For Beginners
A validation curve plots model performance (y-axis) against a hyperparameter value (x-axis). For example, plotting accuracy vs. regularization strength helps you find the best regularization value. The point where train and validation scores are both high and close together is often optimal.
How It Works
Validation curves show how model performance changes as a hyperparameter varies. This helps find optimal hyperparameter values and diagnose overfitting.
Constructors
| Constructor | Summary |
|---|---|
ValidationCurveOptions() |
Properties
| Property | Summary |
|---|---|
CrossValidationStrategy? CVStrategy { get; set; } |
Cross-validation strategy. |
IReadOnlyList<string> MetricsToTrack { get; set; } |
Metrics to track. |
OptimalValueSelectionMethod? OptimalValueMethod { get; set; } |
How to select optimal value. |
bool? ComputeConfidenceIntervals { get; set; } |
Whether to compute confidence intervals. |
bool? FindOptimalValue { get; set; } |
Whether to find optimal parameter value. |
bool? HigherValueMoreComplex { get; set; } |
Whether higher parameter values mean more model complexity. |
bool? ParallelExecution { get; set; } |
Whether to run in parallel. |
bool? Shuffle { get; set; } |
Whether to shuffle before CV. |
bool? UseLogScale { get; set; } |
Whether to use logarithmic spacing for parameter values. |
double? ConfidenceLevel { get; set; } |
Confidence level. |
double[] ParameterValues { get; set; } |
Values to test for the parameter. |
int? CVFolds { get; set; } |
Number of CV folds. |
int? MaxDegreeOfParallelism { get; set; } |
Maximum parallelism. |
int? RandomSeed { get; set; } |
Random seed for reproducibility. |
string ParameterName { get; set; } |
Name of the parameter to vary. |
string ParameterSetterMethod { get; set; } |
Custom parameter setter function name. |