C# Class Encog.Normalize.Segregate.IntegerBalanceSegregator

Balance based on an input value. This allows you to make sure that one input class does not saturate the training data. To do this, you specify the input value to check and the number of occurrences of each integer value of this field to allow.

Inheritance: ISegregator

Show file Open project: encog/encog-silverlight-core

Public Methods

Method	Description
DumpCounts ( ) : String	Get information on how many rows fall into each group.
Init ( DataNormalization normalization ) : void	Init the segregator with the owning normalization object.
IntegerBalanceSegregator ( ) : System	Default constructor for reflection.
IntegerBalanceSegregator ( IInputField target, int count ) : System	Construct a balanced segregator.
PassInit ( ) : void	Init for a new pass.
ShouldInclude ( ) : bool	Determine of the current row should be included.

Method Details

DumpCounts() public method

Get information on how many rows fall into each group.

public DumpCounts ( ) : String
return	String

Init() public method

Init the segregator with the owning normalization object.

public Init ( DataNormalization normalization ) : void
normalization	DataNormalization	The data normalization object to use.
return	void

IntegerBalanceSegregator() public method

Default constructor for reflection.

public IntegerBalanceSegregator ( ) : System
return	System

IntegerBalanceSegregator() public method

Construct a balanced segregator.

public IntegerBalanceSegregator ( IInputField target, int count ) : System
target	IInputField	The input field to base this on, should /// be an integer value.
count	int	The number of rows to accept from each /// unique value for the input.
return	System

PassInit() public method

Init for a new pass.

public PassInit ( ) : void
return	void

ShouldInclude() public method

Determine of the current row should be included.

public ShouldInclude ( ) : bool
return	bool