Examples of HybridStrategy

kodkod.engine.ucore.HybridStrategy
A hybrid strategy for generating unsat cores that are minimal when mapped back onto the logic level. Specifically, let C be a core that is minimal according to this strategy, and let F(C) be the top-level logic constraints corresponding to C. Then, this strategy guarantees that there is no clause c in C such that F(C - c) is a strict subset of F(C). Note that this does not guarantee that F(C) itself is minimal. In other words, there could be an f in F(C) such that F(C) - f is unsatisfiable. To get a minimal logic core, use {@linkplain RCEStrategy}.
This strategy will work properly only on CNFs generated by the kodkod {@linkplain Translator}.
@author Emina Torlak @see RCEStrategy
org.encog.ml.train.strategy.HybridStrategy
A hybrid stragey allows a secondary training algorithm to be used. Once the primary algorithm is no longer improving by much, the secondary will be used. Using simulated annealing in as a secondary to one of the propagation methods is often a very efficient combination as it can help the propagation method escape a local minimum. This is particularly true with backpropagation. @author jheaton

Examples of kodkod.engine.ucore.HybridStrategy

              if (sol.outcome()==UNSATISFIABLE) {
                 // only perform the minimization if it was UNSATISFIABLE, rather than TRIVIALLY_UNSATISFIABLE
                 int i = p.highLevelCore().size();
                 rep.minimizing(cmd, i);
                 if (opt.coreMinimization==0) try { p.minimize(new RCEStrategy(p.log())); } catch(Throwable ex) {}
                 if (opt.coreMinimization==1) try { p.minimize(new HybridStrategy(p.log())); } catch(Throwable ex) {}
                 rep.minimized(cmd, i, p.highLevelCore().size());
              }
              for(Iterator<TranslationRecord> it=p.core(); it.hasNext();) {
                 Object n=it.next().node();
                 if (n instanceof Formula) lCore.add((Formula)n);

View Full Code Here

Examples of org.encog.ml.train.strategy.HybridStrategy

    final MLTrain trainMain = new Backpropagation(network, trainingSet,0.000001, 0.0);


    ((Propagation)trainMain).setNumThreads(1);
    final StopTrainingStrategy stop = new StopTrainingStrategy();
    trainMain.addStrategy(new Greedy());
    trainMain.addStrategy(new HybridStrategy(trainAlt));
    trainMain.addStrategy(stop);


    int epoch = 0;
    while (!stop.shouldStop()) {
      trainMain.iteration();

View Full Code Here

Examples of org.encog.ml.train.strategy.HybridStrategy

    final MLTrain trainMain = new Backpropagation(network, trainingSet,0.000001, 0.0);


    ((Propagation)trainMain).setNumThreads(1);
    final StopTrainingStrategy stop = new StopTrainingStrategy();
    trainMain.addStrategy(new Greedy());
    trainMain.addStrategy(new HybridStrategy(trainAlt));
    trainMain.addStrategy(stop);


    int epoch = 0;
    while (!stop.shouldStop()) {
      trainMain.iteration();

View Full Code Here

Examples of org.encog.ml.train.strategy.HybridStrategy

  
  public void testHybrid()
  {
    MockTrain alt = new MockTrain();
    
    HybridStrategy strategy = new HybridStrategy(alt,0.001,2,5 );
    
    MockTrain mock = new MockTrain();
    mock.addStrategy(strategy);
    mock.setError(0.05);
    mock.iteration();

View Full Code Here

Examples of org.encog.ml.train.strategy.HybridStrategy

  
  public void testHybrid()
  {
    MockTrain alt = new MockTrain();
    
    HybridStrategy strategy = new HybridStrategy(alt,0.001,2,5 );
    
    MockTrain mock = new MockTrain();
    mock.addStrategy(strategy);
    mock.setError(0.05);
    mock.iteration();

View Full Code Here

TOP

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.