Changeset 5816

Timestamp:

03/23/11 12:54:57 (13 years ago)

Author:

gkronber

Message:

#1450 Added preliminary implementation for solution ensemble support.

Location:

branches/DataAnalysis SolutionEnsembles

Files:

• HeuristicLab.Algorithms.DataAnalysis/3.4/CrossValidation.cs (modified) (1 diff)
• HeuristicLab.Problems.DataAnalysis/3.4/HeuristicLab.Problems.DataAnalysis-3.4.csproj (modified) (2 diffs)
• HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Classification/ClassificationEnsembleSolution.cs (added)
• HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Classification/ClassificationSolution.cs (modified) (2 diffs)
• HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Clustering/ClusteringSolution.cs (modified) (2 diffs)
• HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression/RegressionEnsembleSolution.cs (added)
• HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression/RegressionSolution.cs (modified) (2 diffs)
• HeuristicLab.Problems.DataAnalysis/3.4/Interfaces/Classification/IClassificationEnsembleSolution.cs (added)
• HeuristicLab.Problems.DataAnalysis/3.4/Interfaces/Regression/IRegressionEnsembleSolution.cs (added)

branches/DataAnalysis SolutionEnsembles/HeuristicLab.Algorithms.DataAnalysis/3.4/CrossValidation.cs

@@ -355 +355 @@
       foreach (IResult result in ExtractAndAggregateResults<PercentValue>(resultCollections))
         results.Add(result.Name, result.Value);
-
+      foreach (IResult result in ExtractAndAggregateRegressionSolutions(resultCollections)) {
+        results.Add(result.Name, result.Value);
+      }
+      //foreach (IResult result in ExtractAndAggregateClassificationSolutions(resultCollections)) {
+      //  results.Add(result.Name, result.Value);
+      //}
       results.Add("Execution Time", new TimeSpanValue(this.ExecutionTime));
       results.Add("CrossValidation Folds", new RunCollection(runs));
     }
+
+    private IEnumerable<IResult> ExtractAndAggregateRegressionSolutions(IEnumerable<KeyValuePair<string, IItem>> resultCollections) {
+      Dictionary<string, List<IRegressionSolution>> resultSolutions = new Dictionary<string, List<IRegressionSolution>>();
+      foreach (var result in resultCollections) {
+        var regressionSolution = result.Value as IRegressionSolution;
+        if (regressionSolution != null) {
+          if (resultSolutions.ContainsKey(result.Key)) {
+            resultSolutions[result.Key].Add(regressionSolution);
+          } else {
+            resultSolutions.Add(result.Key, new List<IRegressionSolution>() { regressionSolution });
+          }
+        }
+      }
+      List<IResult> aggregatedResults = new List<IResult>();
+      foreach (KeyValuePair<string, List<IRegressionSolution>> solutions in resultSolutions) {
+        var problemDataClone = (IRegressionProblemData)Problem.ProblemData.Clone();
+        problemDataClone.TrainingPartition.Start = SamplesStart.Value; problemDataClone.TrainingPartition.End = SamplesEnd.Value;
+        problemDataClone.TestPartition.Start = SamplesStart.Value; problemDataClone.TestPartition.End = SamplesEnd.Value;
+        var ensembleSolution = new RegressionEnsembleSolution(solutions.Value.Select(x => x.Model), problemDataClone,
+          solutions.Value.Select(x => x.ProblemData.TrainingPartition),
+          solutions.Value.Select(x => x.ProblemData.TestPartition)); ;
+
+        aggregatedResults.Add(new Result(solutions.Key, ensembleSolution));
+      }
+      return aggregatedResults;
+    }
+
+    //private IEnumerable<IResult> ExtractAndAggregateClassificationSolutions(IEnumerable<KeyValuePair<string, IItem>> resultCollections) {
+    //  Dictionary<string, List<IClassificationSolution>> resultSolutions = new Dictionary<string, List<IClassificationSolution>>();
+    //  foreach (var result in resultCollections) {
+    //    var classificationSolution = result.Value as IClassificationSolution;
+    //    if (classificationSolution != null) {
+    //      if (resultSolutions.ContainsKey(result.Key)) {
+    //        resultSolutions[result.Key].Add(classificationSolution);
+    //      } else {
+    //        resultSolutions.Add(result.Key, new List<IClassificationSolution>() { classificationSolution });
+    //      }
+    //    }
+    //  }
+    //  List<IResult> aggregatedResults = new List<IResult>();
+    //  foreach (KeyValuePair<string, List<IClassificationSolution>> solutions in resultSolutions) {
+    //    var ensembleModel = new ClassificationEnsembleModel(solutions.Value.Select(x => x.Model));
+    //    var problemDataClone = (IClassificationProblemData)Problem.ProblemData.Clone();
+
+    //    problemDataClone.TrainingPartition.Start = SamplesStart.Value; problemDataClone.TrainingPartition.End = SamplesEnd.Value;
+    //    problemDataClone.TestPartition.Start = SamplesStart.Value; problemDataClone.TestPartition.End = SamplesEnd.Value;
+    //    aggregatedResults.Add(new Result(solutions.Key, new ClassificationSolution(ensembleModel, problemDataClone)));
+    //  }
+    //  return aggregatedResults;
+    //}
 
     private static IEnumerable<IResult> ExtractAndAggregateResults<T>(IEnumerable<KeyValuePair<string, IItem>> results)

branches/DataAnalysis SolutionEnsembles/HeuristicLab.Problems.DataAnalysis/3.4/HeuristicLab.Problems.DataAnalysis-3.4.csproj

@@ -109 +109 @@
   <ItemGroup>
     <Compile Include="DoubleLimit.cs" />
+    <Compile Include="Implementation\Classification\ClassificationEnsembleModel.cs">
+      <SubType>Code</SubType>
+    </Compile>
     <Compile Include="Implementation\Classification\ClassificationProblemData.cs" />
     <Compile Include="Implementation\Classification\ClassificationProblem.cs" />
@@ -115 +118 @@
     <Compile Include="Implementation\Clustering\ClusteringProblemData.cs" />
     <Compile Include="Implementation\Clustering\ClusteringSolution.cs" />
-    <Compile Include="Implementation\Classification\ClassificationEnsembleModel.cs" />
+    <Compile Include="Implementation\Regression\RegressionEnsembleModel.cs">
+      <SubType>Code</SubType>
+    </Compile>
+    <Compile Include="Implementation\Regression\RegressionEnsembleSolution.cs" />
+    <Compile Include="Interfaces\Classification\IClassificationEnsembleModel.cs">
+      <SubType>Code</SubType>
+    </Compile>
+    <Compile Include="Interfaces\Classification\IClassificationEnsembleSolution.cs" />
     <Compile Include="Interfaces\Classification\IDiscriminantFunctionThresholdCalculator.cs" />
-    <Compile Include="Interfaces\Classification\IClassificationEnsembleModel.cs" />
-    <Compile Include="Interfaces\Regression\IRegressionEnsembleModel.cs" />
+    <Compile Include="Interfaces\Regression\IRegressionEnsembleModel.cs">
+      <SubType>Code</SubType>
+    </Compile>
+    <Compile Include="Interfaces\Regression\IRegressionEnsembleSolution.cs" />
     <Compile Include="OnlineEvaluators\OnlineLinearScalingParameterCalculator.cs" />
-    <Compile Include="Implementation\Regression\RegressionEnsembleModel.cs" />
     <Compile Include="Implementation\Classification\DiscriminantFunctionClassificationModel.cs" />
     <Compile Include="Implementation\Classification\DiscriminantFunctionClassificationSolution.cs" />

branches/DataAnalysis SolutionEnsembles/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Classification/ClassificationSolution.cs

@@ -30 +30 @@
 namespace HeuristicLab.Problems.DataAnalysis {
   /// <summary>
-  /// Abstract base class for classification data analysis solutions
+  /// Represents a classification data analysis solution
   /// </summary>
   [StorableClass]
-  public abstract class ClassificationSolution : DataAnalysisSolution, IClassificationSolution {
+  public class ClassificationSolution : DataAnalysisSolution, IClassificationSolution {
     private const string TrainingAccuracyResultName = "Accuracy (training)";
     private const string TestAccuracyResultName = "Accuracy (test)";
@@ -67 +67 @@
       Add(new Result(TestAccuracyResultName, "Accuracy of the model on the test partition (percentage of correctly classified instances).", new PercentValue()));
       RecalculateResults();
+    }
+
+    public override IDeepCloneable Clone(Cloner cloner) {
+      return new ClassificationSolution(this, cloner);
     }
 

branches/DataAnalysis SolutionEnsembles/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Clustering/ClusteringSolution.cs

@@ -27 +27 @@
 namespace HeuristicLab.Problems.DataAnalysis {
   /// <summary>
-  /// Abstract base class for clustering data analysis solutions
+  /// Represents a clustering data analysis solution
   /// </summary>
   [StorableClass]
-  public abstract class ClusteringSolution : DataAnalysisSolution, IClusteringSolution {
+  public class ClusteringSolution : DataAnalysisSolution, IClusteringSolution {
 
     [StorableConstructor]
@@ -39 +39 @@
     public ClusteringSolution(IClusteringModel model, IClusteringProblemData problemData)
       : base(model, problemData) {
+    }
+
+    public override IDeepCloneable Clone(Cloner cloner) {
+      return new ClusteringSolution(this, cloner);
     }
 

branches/DataAnalysis SolutionEnsembles/HeuristicLab.Problems.DataAnalysis/3.4/Implementation/Regression/RegressionSolution.cs

@@ -30 +30 @@
 namespace HeuristicLab.Problems.DataAnalysis {
   /// <summary>
-  /// Abstract base class for regression data analysis solutions
+  /// Represents a regression data analysis solution
   /// </summary>
   [StorableClass]
-  public abstract class RegressionSolution : DataAnalysisSolution, IRegressionSolution {
+  public class RegressionSolution : DataAnalysisSolution, IRegressionSolution {
     private const string TrainingMeanSquaredErrorResultName = "Mean squared error (training)";
     private const string TestMeanSquaredErrorResultName = "Mean squared error (test)";
@@ -99 +99 @@
     }
 
+    public override IDeepCloneable Clone(Cloner cloner) {
+      return new RegressionSolution(this, cloner);
+    }
+
     protected override void OnProblemDataChanged(EventArgs e) {
       base.OnProblemDataChanged(e);