Examples of joshua.discriminative.training.risk_annealer.hypergraph.FeatureForest

Package joshua.discriminative.training.risk_annealer.hypergraph

Examples of joshua.discriminative.training.risk_annealer.hypergraph.FeatureForest

joshua.discriminative.training.risk_annealer.hypergraph.FeatureForest
ideally, we should first process the hypergraph to store risk and feature information, and then using a feature filter to filter unwanted features. Each feature should have a unique feature ID.

       
        
    for(int sentID=0; sentID < numSentence; sentID ++){
      //System.out.println("#Process sentence " + sent_id);
      
      FeatureForest fForest = hgFactory.nextHG(sentID);
      
      fForest.setFeatureWeights(weightsForTheta);
      fForest.setScale(scalingFactor);
  


      /** Based on a model and a test hypergraph (which provides the topology and feature/risk annotation), 
       *  compute the gradient.
       **/

View Full Code Here

  }


  @Override
  public void consume(HGAndReferences hgAndRefs) {
    
    FeatureForest fForest = riskAnnotator.riskAnnotationOnHG(hgAndRefs.hg, hgAndRefs.referenceSentences);
    
    fForest.setFeatureWeights(weightsForTheta);
    fForest.setScale(scalingFactor);




    /** Based on a model and a test hypergraph 
     * (which provides the topology and feature/risk annotation), 
     *  compute the gradient and function value.

View Full Code Here

       
        
    for(int sentID=0; sentID < numSentence; sentID ++){
      //System.out.println("#Process sentence " + sent_id);
      
      FeatureForest fForest = hgFactory.nextHG(sentID);
      
      fForest.setFeatureWeights(weightsForTheta);
      fForest.setScale(scalingFactor);
  


      /** Based on a model and a test hypergraph (which provides the topology and feature/risk annotation), 
       *  compute the gradient.
       **/

View Full Code Here

   


   public FeatureForest nextHG(int sentID){
     
    //=== feature forests 
    FeatureForest fForest;


    if( saveHGInMemory==false){//on disk
      fForest = readOneHGFromDisk();
    }else{//in memory
      fForest = hyperGraphs.get(sentID);

View Full Code Here

   
   
   private void readHGsIntoMemoryAllAtOnce(int numSent){
    initDiskReading();
    for(int i=0; i<numSent; i++){
      FeatureForest fForest = readOneHGFromDisk();
      hyperGraphs.add(fForest);
    }
    finalizeDiskReading();
   }

View Full Code Here


  
  public FeatureForest riskAnnotationOnHG(HyperGraph hg, String refSentStr){


    setupRefAndPrefixAndSurfixTbl(refSentStr);//TODO: should enforce this in parent class    
    return new FeatureForest(  splitHG(hg) );
  }

View Full Code Here

  }
  
  public FeatureForest riskAnnotationOnHG(HyperGraph hg, String[] refSentStrs){


    setupRefAndPrefixAndSurfixTbl(refSentStrs);//TODO: should enforce this in parent class    
    return new FeatureForest(  splitHG(hg) );
  }

View Full Code Here

TOP

Related Classes of joshua.discriminative.training.risk_annealer.hypergraph.FeatureForest

joshua.discriminative.semiring_parsingv2.pmodule.SparseMap

joshua.discriminative.semiring_parsingv2.SignedValue

joshua.discriminative.training.risk_annealer.hypergraph.deprecated.FeatureForestFactory

joshua.discriminative.training.risk_annealer.hypergraph.deprecated.HGRiskGradientComputerBasedOnSemringV1

joshua.discriminative.training.risk_annealer.hypergraph.deprecated.HGRiskGradientComputerBasedOnSemringV2

joshua.discriminative.training.risk_annealer.hypergraph.deprecated.RiskAndFeatureAnnotation

joshua.discriminative.training.risk_annealer.hypergraph.parallel.GradientConsumer

All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.