C# Class numl.Learner

Primary class for running model generators. It is designed to abstract the separation of training and test sets as well as best model selection.

Afficher le fichier Open project: sethjuarez/numl

Méthodes publiques

Méthode	Description
Best ( this models, ScoringMetric metric = ScoringMetric.Accuracy ) : LearningModel	Retrieve best model (or model with the highest accuracy)
Learn ( IEnumerable	Trains a single model based on a generator a predefined number of times with the provided examples and data split and selects the best (or most accurate) model.
Learn ( IEnumerable	Trains an arbitrary number of models on the provided examples by creating a separation of data based on training percentage. Each generator is rerun a predetermined amount of times.

Private Methods

Méthode	Description
GenerateModel ( IGenerator generator, Matrix x, Vector y, IEnumerable	Generates a model.
GetTestExamples ( IEnumerable slice, IEnumerable	Gets test examples.
GetTestPoints ( int testCount, int total ) : IEnumerable	Gets the test points in this collection.
GetTrainingPoints ( IEnumerable testPoints, int total ) : IEnumerable	Gets the training points in this collection.
Learner ( ) : System	Static constructor.

Method Details

Best() public static méthode

Retrieve best model (or model with the highest accuracy)

public static Best ( this models, ScoringMetric metric = ScoringMetric.Accuracy ) : LearningModel
models	this	List of models.
metric	ScoringMetric	Scoring metric to use for model selection.
Résultat	LearningModel

Learn() public static méthode

Trains a single model based on a generator a predefined number of times with the provided examples and data split and selects the best (or most accurate) model.

public static Learn ( IEnumerable
examples	IEnumerable	Source data.
trainingPercentage	double	Data split percentage.
repeat	int	Number of repetitions per generator.
generator	IGenerator	Model generator used.
Résultat	LearningModel

Learn() public static méthode

Trains an arbitrary number of models on the provided examples by creating a separation of data based on training percentage. Each generator is rerun a predetermined amount of times.

Thrown when the requested operation is invalid.

public static Learn ( IEnumerable
examples	IEnumerable	Source data.
trainingPercentage	double	Data split percentage.
repeat	int	Number of repetitions per generator.
Résultat	numl.LearningModel[]