1 | #region License Information
|
---|
2 | /* HeuristicLab
|
---|
3 | * Copyright (C) 2002-2016 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
4 | *
|
---|
5 | * This file is part of HeuristicLab.
|
---|
6 | *
|
---|
7 | * HeuristicLab is free software: you can redistribute it and/or modify
|
---|
8 | * it under the terms of the GNU General Public License as published by
|
---|
9 | * the Free Software Foundation, either version 3 of the License, or
|
---|
10 | * (at your option) any later version.
|
---|
11 | *
|
---|
12 | * HeuristicLab is distributed in the hope that it will be useful,
|
---|
13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
15 | * GNU General Public License for more details.
|
---|
16 | *
|
---|
17 | * You should have received a copy of the GNU General Public License
|
---|
18 | * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
|
---|
19 | */
|
---|
20 | #endregion
|
---|
21 |
|
---|
22 | using System;
|
---|
23 | using System.Collections.Generic;
|
---|
24 | using HeuristicLab.Common;
|
---|
25 | using HeuristicLab.Core;
|
---|
26 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
|
---|
27 | using HeuristicLab.Persistence;
|
---|
28 |
|
---|
29 | namespace HeuristicLab.Problems.DataAnalysis.Symbolic.Classification {
|
---|
30 | [StorableType("a9d6ca7c-519e-417a-90ec-f7f42c27820a")]
|
---|
31 | [Item("SymbolicClassificationSolutionImpactValuesCalculator", "Calculate symbolic expression tree node impact values for classification problems.")]
|
---|
32 | public class SymbolicClassificationSolutionImpactValuesCalculator : SymbolicDataAnalysisSolutionImpactValuesCalculator {
|
---|
33 | public SymbolicClassificationSolutionImpactValuesCalculator() { }
|
---|
34 | protected SymbolicClassificationSolutionImpactValuesCalculator(SymbolicClassificationSolutionImpactValuesCalculator original, Cloner cloner)
|
---|
35 | : base(original, cloner) { }
|
---|
36 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
37 | return new SymbolicClassificationSolutionImpactValuesCalculator(this, cloner);
|
---|
38 | }
|
---|
39 | [StorableConstructor]
|
---|
40 | protected SymbolicClassificationSolutionImpactValuesCalculator(StorableConstructorFlag deserializing) : base(deserializing) { }
|
---|
41 |
|
---|
42 | public override void CalculateImpactAndReplacementValues(ISymbolicDataAnalysisModel model,
|
---|
43 | ISymbolicExpressionTreeNode node,
|
---|
44 | IDataAnalysisProblemData problemData, IEnumerable<int> rows, out double impactValue, out double replacementValue,
|
---|
45 | out double newQualityForImpactsCalculation,
|
---|
46 | double qualityForImpactsCalculation = Double.NaN) {
|
---|
47 | var classificationModel = (ISymbolicClassificationModel)model;
|
---|
48 | var classificationProblemData = (IClassificationProblemData)problemData;
|
---|
49 |
|
---|
50 | if (double.IsNaN(qualityForImpactsCalculation))
|
---|
51 | qualityForImpactsCalculation = CalculateQualityForImpacts(classificationModel, classificationProblemData, rows);
|
---|
52 |
|
---|
53 |
|
---|
54 | var cloner = new Cloner();
|
---|
55 | var tempModel = cloner.Clone(classificationModel);
|
---|
56 | var tempModelNode = (ISymbolicExpressionTreeNode)cloner.GetClone(node);
|
---|
57 |
|
---|
58 | var tempModelParentNode = tempModelNode.Parent;
|
---|
59 | int i = tempModelParentNode.IndexOfSubtree(tempModelNode);
|
---|
60 | double bestReplacementValue = 0.0;
|
---|
61 | double bestImpactValue = double.PositiveInfinity;
|
---|
62 | newQualityForImpactsCalculation = qualityForImpactsCalculation; // initialize
|
---|
63 | // try the potentially reasonable replacement values and use the best one
|
---|
64 | foreach (var repValue in CalculateReplacementValues(node, classificationModel.SymbolicExpressionTree, classificationModel.Interpreter, classificationProblemData.Dataset, classificationProblemData.TrainingIndices)) {
|
---|
65 | tempModelParentNode.RemoveSubtree(i);
|
---|
66 |
|
---|
67 | var constantNode = new ConstantTreeNode(new Constant()) { Value = repValue };
|
---|
68 | tempModelParentNode.InsertSubtree(i, constantNode);
|
---|
69 |
|
---|
70 | var dataset = classificationProblemData.Dataset;
|
---|
71 | var targetClassValues = dataset.GetDoubleValues(classificationProblemData.TargetVariable, rows);
|
---|
72 | var estimatedClassValues = tempModel.GetEstimatedClassValues(dataset, rows);
|
---|
73 | OnlineCalculatorError errorState;
|
---|
74 | newQualityForImpactsCalculation = OnlineAccuracyCalculator.Calculate(targetClassValues, estimatedClassValues,
|
---|
75 | out errorState);
|
---|
76 | if (errorState != OnlineCalculatorError.None) newQualityForImpactsCalculation = 0.0;
|
---|
77 |
|
---|
78 | impactValue = qualityForImpactsCalculation - newQualityForImpactsCalculation;
|
---|
79 |
|
---|
80 | if (impactValue < bestImpactValue) {
|
---|
81 | bestImpactValue = impactValue;
|
---|
82 | bestReplacementValue = repValue;
|
---|
83 | }
|
---|
84 | }
|
---|
85 | replacementValue = bestReplacementValue;
|
---|
86 | impactValue = bestImpactValue;
|
---|
87 | }
|
---|
88 |
|
---|
89 | public static double CalculateQualityForImpacts(ISymbolicClassificationModel model, IClassificationProblemData problemData, IEnumerable<int> rows) {
|
---|
90 | OnlineCalculatorError errorState;
|
---|
91 | var dataset = problemData.Dataset;
|
---|
92 | var targetClassValues = dataset.GetDoubleValues(problemData.TargetVariable, rows);
|
---|
93 | var originalClassValues = model.GetEstimatedClassValues(dataset, rows);
|
---|
94 | var qualityForImpactsCalculation = OnlineAccuracyCalculator.Calculate(targetClassValues, originalClassValues, out errorState);
|
---|
95 | if (errorState != OnlineCalculatorError.None) qualityForImpactsCalculation = 0.0;
|
---|
96 |
|
---|
97 | return qualityForImpactsCalculation;
|
---|
98 | }
|
---|
99 | }
|
---|
100 | }
|
---|