com.mancrd.ahah.samples.clusterer.sweep
Class ClustererDriver

java.lang.Object
  extended by com.mancrd.ahah.samples.clusterer.sweep.ClustererDriver

public class ClustererDriver
extends Object

Given a IClusterer and a RandomSpikesGenerator, this class is used to teach and test a Clusterer.

Author:
timmolter

Constructor Summary
ClustererDriver(IClusterer clusterer, RandomSpikePatternGenerator randomSpikesGenerator)
          Constructor
 
Method Summary
 void learnAll(int numTrials)
          Inputs all spikes in the spikes list randomly picked for each trial.
 void learnRandom(int numTrials)
          Inputs only a single spikes in the spikes list randomly picked for each trial.
 VergenceEvaluator testAllSpikes(int numTrials)
          Tests every single spikes in the spikes list for each trial.
 VergenceEvaluator testRandomSpikes(int numTrials)
          Tests only a single spikes in the spikes list randomly picked for each trial.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

ClustererDriver

public ClustererDriver(IClusterer clusterer,
                       RandomSpikePatternGenerator randomSpikesGenerator)
Constructor

Method Detail

learnRandom

public void learnRandom(int numTrials)
Inputs only a single spikes in the spikes list randomly picked for each trial. Features are cloned with noise. Total inputs will be N, where N=numTrials

Parameters:
numTrials -

learnAll

public void learnAll(int numTrials)
Inputs all spikes in the spikes list randomly picked for each trial. Features are cloned with noise. Total inputs will be N*numSpikes, where N=numTrials

Parameters:
numTrials -

testAllSpikes

public VergenceEvaluator testAllSpikes(int numTrials)
Tests every single spikes in the spikes list for each trial. Features are cloned with noise. Total inputs will be M*N, where M=numFeatures and N=numTrials

Parameters:
numTrials -

testRandomSpikes

public VergenceEvaluator testRandomSpikes(int numTrials)
Tests only a single spikes in the spikes list randomly picked for each trial. Features are cloned with noise. Total inputs will be N, where N=numTrials

Parameters:
numTrials -


Copyright © 2013–2014 M. Alexander Nugent Consulting, Research and Devlopment. All rights reserved.