org.encog.ml.data
Interface MLSequenceSet

All Superinterfaces:
Iterable<MLDataPair>, MLDataSet
All Known Implementing Classes:
BasicMLSequenceSet

public interface MLSequenceSet
extends MLDataSet

A sequence set is a collection of data sets. Where each individual data set is one "unbroken sequence" within the sequence set. This allows individual observations to occur individually, indicating a break between them. The sequence set, itself, is a data set, so it can be used with any Encog trainer. However, not all trainers are aware of sequence sets. Further, some machine learning methods are unaffected by them. Sequence sets are typically used with Hidden Markov Models (HMM)'s.


Method Summary
 void add(MLDataSet sequence)
          Add a new sequence.
 MLDataSet getSequence(int i)
          Get an individual sequence.
 int getSequenceCount()
           
 Collection<MLDataSet> getSequences()
           
 void startNewSequence()
          Cause a "break" in the data by creating a the beginning of a new sequence.
 
Methods inherited from interface org.encog.ml.data.MLDataSet
add, add, add, close, get, getIdealSize, getInputSize, getRecord, getRecordCount, isSupervised, openAdditional, size
 
Methods inherited from interface java.lang.Iterable
iterator
 

Method Detail

startNewSequence

void startNewSequence()
Cause a "break" in the data by creating a the beginning of a new sequence.


getSequenceCount

int getSequenceCount()
Returns:
Get a count of the number of sequences.

getSequence

MLDataSet getSequence(int i)
Get an individual sequence.

Parameters:
i - The index of the sequence.
Returns:
The sequence.

getSequences

Collection<MLDataSet> getSequences()
Returns:
A list of all of the sequences.

add

void add(MLDataSet sequence)
Add a new sequence.

Parameters:
sequence - The sequence to add.


Copyright © 2014. All Rights Reserved.