C# Class Accord.MachineLearning.BaseBatches

Utility class for preparing mini-batches of data.

Inheritance: IEnumerable

Datei anzeigen Open project: Webreaper/Damselfly

Public Methods

Method	Description
BaseBatches ( Array inputs, Array weights ) : System	Initializes a new instance of the BaseBatches{TBatch, TInput} class.
GetEnumerator ( ) : IEnumerator	Returns an enumerator that iterates through a collection.

Protected Methods

Method	Description
Init ( ) : TBatch	Inheritors should use this method to create a new instance of a mini-batch object. You can use this method to create mini-batches containing different objects related to the mini-batch, such as auxiliary data, privileged information, etc).
PrepareBatch ( Array idx ) : void	Inheritors should use this method to prepare the data for the next batches, for example by reshuffling according to the ordering passed as argument.
PutCurrentSampleInMiniBatch ( batch, int positionInMiniBatch ) : void	Inheritors should use this method to put the current sample in the given mini-batch at the given position.

Private Methods

Method	Description
IEnumerable ( ) : IEnumerator	Returns an enumerator that iterates through a collection.

Method Details

BaseBatches() public method

Initializes a new instance of the BaseBatches{TBatch, TInput} class.

public BaseBatches ( Array inputs, Array weights ) : System
inputs	Array	The input data that should be divided into batches.
weights	Array	The weight for the data that should be divided into batches.
return	System

GetEnumerator() public method

Returns an enumerator that iterates through a collection.

public GetEnumerator ( ) : IEnumerator
return	IEnumerator

Init() protected abstract method

Inheritors should use this method to create a new instance of a mini-batch object. You can use this method to create mini-batches containing different objects related to the mini-batch, such as auxiliary data, privileged information, etc).

protected abstract Init ( ) : TBatch
return	TBatch

PrepareBatch() protected method

Inheritors should use this method to prepare the data for the next batches, for example by reshuffling according to the ordering passed as argument.

protected PrepareBatch ( Array idx ) : void
idx	Array	The ordering for the samples in the new batches.
return	void

PutCurrentSampleInMiniBatch() protected method

Inheritors should use this method to put the current sample in the given mini-batch at the given position.

protected PutCurrentSampleInMiniBatch ( batch, int positionInMiniBatch ) : void
batch		The mini-batch being constructed.
positionInMiniBatch	int	The position in mini-batch where the current sample must be put.
return	void