clarion.nacs
Class AMNet

java.lang.Object
  extended byclarion.common.BPNet
      extended byclarion.nacs.AMNet

public class AMNet
extends BPNet


Field Summary
 
Fields inherited from class clarion.common.BPNet
BLAT, BLDT, BLPT, BLRT, debug, DesiredOutput, Errors, Eta, Hidden, HiddenDeriv, HiddenErrors, HiddenMomentum, HiddenThresholds, HiddenToOutputMomentum, HiddenToOutputWeights, HINITTHRESHOLD, HINITWEIGHT, HtoOWeights, Input, InputToHiddenMomentum, InputToHiddenWeights, ItoHWeights, LINITTHRESHOLD, LINITWEIGHT, Momentum, nHidden, nInput, NM, nOutput, Output, OutputDeriv, OutputMomentum, OutputThresholds, PM, RZero, SUCC_C3, SUCC_C4, succRate, sumSqErrors
 
Constructor Summary
AMNet(int nIdx, Global g, GKS gks)
          new constructor, # of input nodes should be equal to # of output nodes.
AMNet(int inputNum, int hiddenNum, int outputNum, int dNum, int[] dValNums)
           
AMNet(int inputNum, int hiddenNum, int outputNum, int dNum, short[] dValNums)
           
 
Method Summary
 void getOutput(double[] arr)
          Returns the output in terms of one dimensional array after filtering noise.
 void getOutput(double[][] arr)
          Returns the output in terms of two dimensional array after filtering noise.
 Feature[] getOutputFeatures(boolean multiFlag)
          Returns the output in terms of features.
 Feature[] getResults()
          Returns the retrieval results in terms of features.
 int getRT()
          Returns the response time.
 void reinit()
          reinitialization routine.
 void resetRT()
          reset the response time.
 boolean retrieval(boolean multiFlag)
          the retrieval process.
 void setAssocAssimOn(boolean flag)
          Sets the flag for assimulation or not.
 void setDesiredOutput(Feature[] features)
          Sets the desired output with the specific features.
 void setDesiredOutput(short[] arr)
          Deprecated. Replcaced by setOutput( short[][] arr ).
 void setDesiredOutput(short[][] arr)
          Sets the desired output with the specific info of active values of a chunk in terms of two dimensional array.
 void setInput(double[][] arr)
          Sets the input with the specific info of activation of a chunk in terms of two dimensional array.
 void setInput(Feature[] features)
          Sets the input with the specific features.
 void setInput(int[] arr)
          Deprecated. Replcaced by setInput( short[][] arr ).
 void setInput(short[][] arr)
          Sets the input with the specific info of active values of a chunk in terms of two dimensional array.
 void setInput(short[][][] arr)
          Sets the input with the specific info of active values of chunks in terms of three dimensional array.
 void setNoiseThreshold(double th)
          Sets noise threshold.
 void setPassNum(int num)
           
 void training()
          the assimulation (training) process.
 void update()
          update the AMN.
 
Methods inherited from class clarion.common.BPNet
backwardPass, calcRT, computeErrors, computeHiddenActivation, computeHiddenErrors, computeOutputActivation, forwardPass, getInput, getnHidden, getnInput, getNM, getnOutput, getOutput, getOutput, getOutput, getOutput, getPM, getResponseTime, getSuccRate, getSumSqErrors, modifyHiddenToOutput, modifyInputToHidden, reinitWeights, resetMatches, restoreInitWeights, setDesiredOutput, setDesiredOutput, setInput, setInput, setLearningRate, setMomentum, updateMatches
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

AMNet

public AMNet(int inputNum,
             int hiddenNum,
             int outputNum,
             int dNum,
             int[] dValNums)

AMNet

public AMNet(int inputNum,
             int hiddenNum,
             int outputNum,
             int dNum,
             short[] dValNums)

AMNet

public AMNet(int nIdx,
             Global g,
             GKS gks)
new constructor, # of input nodes should be equal to # of output nodes.

Parameters:
nIdx - the AMN index.
g - the associated global settings.
gks - the associated GKS.
Method Detail

reinit

public void reinit()
reinitialization routine.


getResults

public Feature[] getResults()
Returns the retrieval results in terms of features.

Returns:
the retrieval results.

getOutput

public void getOutput(double[] arr)
Returns the output in terms of one dimensional array after filtering noise.

Overrides:
getOutput in class BPNet
Parameters:
arr - the array to store current output.
Returns:
the retrieval results.

getOutput

public void getOutput(double[][] arr)
Returns the output in terms of two dimensional array after filtering noise.

Returns:
the retrieval results.

getOutputFeatures

public Feature[] getOutputFeatures(boolean multiFlag)
Returns the output in terms of features.

Returns:
the output features.

resetRT

public void resetRT()
reset the response time.


getRT

public int getRT()
Returns the response time.

Returns:
the response time.

setNoiseThreshold

public void setNoiseThreshold(double th)
Sets noise threshold.

Parameters:
th - the value to set the noise threshold.

setAssocAssimOn

public void setAssocAssimOn(boolean flag)
Sets the flag for assimulation or not.


setPassNum

public void setPassNum(int num)

setInput

public void setInput(int[] arr)
Deprecated. Replcaced by setInput( short[][] arr ).

Sets the input with the specific info of active values of a chunk in terms of one dimensional array.

Overrides:
setInput in class BPNet
Parameters:
arr - the content used as input.

setInput

public void setInput(short[][] arr)
Sets the input with the specific info of active values of a chunk in terms of two dimensional array.

Parameters:
arr - the content used as input.

setInput

public void setInput(double[][] arr)
Sets the input with the specific info of activation of a chunk in terms of two dimensional array.

Parameters:
arr - the content used as input.

setInput

public void setInput(short[][][] arr)
Sets the input with the specific info of active values of chunks in terms of three dimensional array.

Parameters:
arr - the content used as input.

setInput

public void setInput(Feature[] features)
Sets the input with the specific features.

Parameters:
features - the features used as input.

setDesiredOutput

public void setDesiredOutput(short[] arr)
Deprecated. Replcaced by setOutput( short[][] arr ).

Sets the desired output with the specific info of active values of a chunk in terms of one dimensional array.

Overrides:
setDesiredOutput in class BPNet
Parameters:
arr - the content used as output.

setDesiredOutput

public void setDesiredOutput(short[][] arr)
Sets the desired output with the specific info of active values of a chunk in terms of two dimensional array.

Parameters:
arr - the content used as input.

setDesiredOutput

public void setDesiredOutput(Feature[] features)
Sets the desired output with the specific features.

Overrides:
setDesiredOutput in class BPNet
Parameters:
features - the features used as input.

training

public void training()
the assimulation (training) process.


retrieval

public boolean retrieval(boolean multiFlag)
the retrieval process.

Parameters:
multiFlag - allow multiple active values in one dimension.

update

public void update()
update the AMN.