C# Class MyMediaLite.Classification.LogisticRegression

Regularized logistic regression trained by stochastic gradient descent

Implementation for dense feature vectors. Predictor variables are implemented transposed to their normal layout, due to being used in for the KDD Cup 2011 ensembles.

显示文件 Open project: zenogantner/MML-KDD Class Usage Examples

Public Properties

Property	Type	Description
parameters	IList

Public Methods

Method	Description
InitModel ( ) : void
LogisticRegression ( ) : System	Default constructor
PredictProbability ( IList input ) : double	Predict probability for given features
Train ( ) : void	Train using stochastic gradient descent

Method Details

InitModel() public method

public InitModel ( ) : void
return	void

LogisticRegression() public method

Default constructor

public LogisticRegression ( ) : System
return	System

PredictProbability() public method

Predict probability for given features

public PredictProbability ( IList input ) : double
input	IList	the input
return	double

Train() public method

Train using stochastic gradient descent

public Train ( ) : void
return	void

Property Details

parameters public_oe property

public IList parameters
return	IList