Changeset 12290 for branches/HeuristicLab.Problems.GrammaticalOptimization-gkr/HeuristicLab.Algorithms.Bandits/Policies/BoltzmannExplorationPolicy.cs

Timestamp:

04/07/15 14:31:06 (10 years ago)

Author:

gkronber

Message:

#2283 created a new branch to separate development from aballeit

Location:

branches/HeuristicLab.Problems.GrammaticalOptimization-gkr

Files:

• . (copied) (copied from branches/HeuristicLab.Problems.GrammaticalOptimization)
• HeuristicLab.Algorithms.Bandits/Policies/BoltzmannExplorationPolicy.cs (modified) (2 diffs)

branches/HeuristicLab.Problems.GrammaticalOptimization-gkr/HeuristicLab.Algorithms.Bandits/Policies/BoltzmannExplorationPolicy.cs

@@ -11 +11 @@
   public class BoltzmannExplorationPolicy : IBanditPolicy {
     private readonly double beta;
-    private readonly Func<DefaultPolicyActionInfo, double> valueFunction;
 
-    public BoltzmannExplorationPolicy(double beta) : this(beta, DefaultPolicyActionInfo.AverageReward) { }
-
-    public BoltzmannExplorationPolicy(double beta, Func<DefaultPolicyActionInfo, double> valueFunction) {
+    public BoltzmannExplorationPolicy(double beta)  { 
       if (beta < 0) throw new ArgumentException();
       this.beta = beta;
-      this.valueFunction = valueFunction;
     }
     public int SelectAction(Random random, IEnumerable<IBanditPolicyActionInfo> actionInfos) {
@@ -37 +33 @@
 
       var w = from aInfo in myActionInfos
-              select Math.Exp(beta * valueFunction(aInfo));
+              select Math.Exp(beta * aInfo.Value);
 
       var bestAction = Enumerable.Range(0, myActionInfos.Count()).SampleProportional(random, w);