| [6589] | 1 | #region License Information
|
|---|
| 2 | /* HeuristicLab
|
|---|
| [11171] | 3 | * Copyright (C) 2002-2014 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
|---|
| [6589] | 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.Collections.Generic;
|
|---|
| 23 | using System.Linq;
|
|---|
| 24 | using HeuristicLab.Common;
|
|---|
| 25 | using HeuristicLab.Data;
|
|---|
| 26 | using HeuristicLab.Optimization;
|
|---|
| 27 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
|---|
| 28 |
|
|---|
| 29 | namespace HeuristicLab.Problems.DataAnalysis {
|
|---|
| 30 | [StorableClass]
|
|---|
| 31 | public abstract class ClassificationSolutionBase : DataAnalysisSolution, IClassificationSolution {
|
|---|
| 32 | private const string TrainingAccuracyResultName = "Accuracy (training)";
|
|---|
| 33 | private const string TestAccuracyResultName = "Accuracy (test)";
|
|---|
| [6913] | 34 | private const string TrainingNormalizedGiniCoefficientResultName = "Normalized Gini Coefficient (training)";
|
|---|
| 35 | private const string TestNormalizedGiniCoefficientResultName = "Normalized Gini Coefficient (test)";
|
|---|
| [6589] | 36 |
|
|---|
| [11622] | 37 | private const string TrainingTruePositiveRateResultName = "True positive rate (training)";
|
|---|
| 38 | private const string TrainingTrueNegativeRateResultName = "True negative rate (training)";
|
|---|
| 39 | private const string TrainingPositivePredictiveValueResultName = "Positive predictive value (training)";
|
|---|
| 40 | private const string TrainingNegativePredictiveValueResultName = "Negative predictive value (training)";
|
|---|
| 41 | private const string TrainingFalsePositiveRateResultName = "False positive rate (training)";
|
|---|
| 42 | private const string TrainingFalseDiscoveryRateResultName = "False discovery rate (training)";
|
|---|
| 43 | private const string TestTruePositiveRateResultName = "True positive rate (test)";
|
|---|
| 44 | private const string TestTrueNegativeRateResultName = "True negative rate (test)";
|
|---|
| 45 | private const string TestPositivePredictiveValueResultName = "Positive predictive value (test)";
|
|---|
| 46 | private const string TestNegativePredictiveValueResultName = "Negative predictive value (test)";
|
|---|
| 47 | private const string TestFalsePositiveRateResultName = "False positive rate (test)";
|
|---|
| 48 | private const string TestFalseDiscoveryRateResultName = "False discovery rate (test)";
|
|---|
| 49 | private const string QualityMeasuresResultName = "Classification Quality Measures";
|
|---|
| 50 |
|
|---|
| [6589] | 51 | public new IClassificationModel Model {
|
|---|
| 52 | get { return (IClassificationModel)base.Model; }
|
|---|
| 53 | protected set { base.Model = value; }
|
|---|
| 54 | }
|
|---|
| 55 |
|
|---|
| 56 | public new IClassificationProblemData ProblemData {
|
|---|
| 57 | get { return (IClassificationProblemData)base.ProblemData; }
|
|---|
| [6653] | 58 | set { base.ProblemData = value; }
|
|---|
| [6589] | 59 | }
|
|---|
| 60 |
|
|---|
| 61 | #region Results
|
|---|
| 62 | public double TrainingAccuracy {
|
|---|
| 63 | get { return ((DoubleValue)this[TrainingAccuracyResultName].Value).Value; }
|
|---|
| 64 | private set { ((DoubleValue)this[TrainingAccuracyResultName].Value).Value = value; }
|
|---|
| 65 | }
|
|---|
| 66 | public double TestAccuracy {
|
|---|
| 67 | get { return ((DoubleValue)this[TestAccuracyResultName].Value).Value; }
|
|---|
| 68 | private set { ((DoubleValue)this[TestAccuracyResultName].Value).Value = value; }
|
|---|
| 69 | }
|
|---|
| [6913] | 70 | public double TrainingNormalizedGiniCoefficient {
|
|---|
| 71 | get { return ((DoubleValue)this[TrainingNormalizedGiniCoefficientResultName].Value).Value; }
|
|---|
| 72 | protected set { ((DoubleValue)this[TrainingNormalizedGiniCoefficientResultName].Value).Value = value; }
|
|---|
| 73 | }
|
|---|
| 74 | public double TestNormalizedGiniCoefficient {
|
|---|
| 75 | get { return ((DoubleValue)this[TestNormalizedGiniCoefficientResultName].Value).Value; }
|
|---|
| 76 | protected set { ((DoubleValue)this[TestNormalizedGiniCoefficientResultName].Value).Value = value; }
|
|---|
| 77 | }
|
|---|
| [11622] | 78 |
|
|---|
| 79 | #region Quality Measures
|
|---|
| 80 | public ResultCollection QualityMeasures {
|
|---|
| 81 | get { return ((ResultCollection)this[QualityMeasuresResultName].Value); }
|
|---|
| 82 | protected set { (this[QualityMeasuresResultName].Value) = value; }
|
|---|
| 83 | }
|
|---|
| 84 |
|
|---|
| 85 | public double TrainingTruePositiveRate {
|
|---|
| 86 | get { return ((DoubleValue)QualityMeasures[TrainingTruePositiveRateResultName].Value).Value; }
|
|---|
| 87 | protected set { ((DoubleValue)QualityMeasures[TrainingTruePositiveRateResultName].Value).Value = value; }
|
|---|
| 88 | }
|
|---|
| 89 | public double TrainingTrueNegativeRate {
|
|---|
| 90 | get { return ((DoubleValue)QualityMeasures[TrainingTrueNegativeRateResultName].Value).Value; }
|
|---|
| 91 | protected set { ((DoubleValue)QualityMeasures[TrainingTrueNegativeRateResultName].Value).Value = value; }
|
|---|
| 92 | }
|
|---|
| 93 | public double TrainingPositivePredictiveValue {
|
|---|
| 94 | get { return ((DoubleValue)QualityMeasures[TrainingPositivePredictiveValueResultName].Value).Value; }
|
|---|
| 95 | protected set { ((DoubleValue)QualityMeasures[TrainingPositivePredictiveValueResultName].Value).Value = value; }
|
|---|
| 96 | }
|
|---|
| 97 | public double TrainingNegativePredictiveValue {
|
|---|
| 98 | get { return ((DoubleValue)QualityMeasures[TrainingNegativePredictiveValueResultName].Value).Value; }
|
|---|
| 99 | protected set { ((DoubleValue)QualityMeasures[TrainingNegativePredictiveValueResultName].Value).Value = value; }
|
|---|
| 100 | }
|
|---|
| 101 | public double TrainingFalsePositiveRate {
|
|---|
| 102 | get { return ((DoubleValue)QualityMeasures[TrainingFalsePositiveRateResultName].Value).Value; }
|
|---|
| 103 | protected set { ((DoubleValue)QualityMeasures[TrainingFalsePositiveRateResultName].Value).Value = value; }
|
|---|
| 104 | }
|
|---|
| 105 | public double TrainingFalseDiscoveryRate {
|
|---|
| 106 | get { return ((DoubleValue)QualityMeasures[TrainingFalseDiscoveryRateResultName].Value).Value; }
|
|---|
| 107 | protected set { ((DoubleValue)QualityMeasures[TrainingFalseDiscoveryRateResultName].Value).Value = value; }
|
|---|
| 108 | }
|
|---|
| 109 |
|
|---|
| 110 | public double TestTruePositiveRate {
|
|---|
| 111 | get { return ((DoubleValue)QualityMeasures[TestTruePositiveRateResultName].Value).Value; }
|
|---|
| 112 | protected set { ((DoubleValue)QualityMeasures[TestTruePositiveRateResultName].Value).Value = value; }
|
|---|
| 113 | }
|
|---|
| 114 | public double TestTrueNegativeRate {
|
|---|
| 115 | get { return ((DoubleValue)QualityMeasures[TestTrueNegativeRateResultName].Value).Value; }
|
|---|
| 116 | protected set { ((DoubleValue)QualityMeasures[TestTrueNegativeRateResultName].Value).Value = value; }
|
|---|
| 117 | }
|
|---|
| 118 | public double TestPositivePredictiveValue {
|
|---|
| 119 | get { return ((DoubleValue)QualityMeasures[TestPositivePredictiveValueResultName].Value).Value; }
|
|---|
| 120 | protected set { ((DoubleValue)QualityMeasures[TestPositivePredictiveValueResultName].Value).Value = value; }
|
|---|
| 121 | }
|
|---|
| 122 | public double TestNegativePredictiveValue {
|
|---|
| 123 | get { return ((DoubleValue)QualityMeasures[TestNegativePredictiveValueResultName].Value).Value; }
|
|---|
| 124 | protected set { ((DoubleValue)QualityMeasures[TestNegativePredictiveValueResultName].Value).Value = value; }
|
|---|
| 125 | }
|
|---|
| 126 | public double TestFalsePositiveRate {
|
|---|
| 127 | get { return ((DoubleValue)QualityMeasures[TestFalsePositiveRateResultName].Value).Value; }
|
|---|
| 128 | protected set { ((DoubleValue)QualityMeasures[TestFalsePositiveRateResultName].Value).Value = value; }
|
|---|
| 129 | }
|
|---|
| 130 | public double TestFalseDiscoveryRate {
|
|---|
| 131 | get { return ((DoubleValue)QualityMeasures[TestFalseDiscoveryRateResultName].Value).Value; }
|
|---|
| 132 | protected set { ((DoubleValue)QualityMeasures[TestFalseDiscoveryRateResultName].Value).Value = value; }
|
|---|
| 133 | }
|
|---|
| [6589] | 134 | #endregion
|
|---|
| [11622] | 135 | #endregion
|
|---|
| [6589] | 136 |
|
|---|
| 137 | [StorableConstructor]
|
|---|
| 138 | protected ClassificationSolutionBase(bool deserializing) : base(deserializing) { }
|
|---|
| 139 | protected ClassificationSolutionBase(ClassificationSolutionBase original, Cloner cloner)
|
|---|
| 140 | : base(original, cloner) {
|
|---|
| 141 | }
|
|---|
| 142 | protected ClassificationSolutionBase(IClassificationModel model, IClassificationProblemData problemData)
|
|---|
| 143 | : base(model, problemData) {
|
|---|
| 144 | Add(new Result(TrainingAccuracyResultName, "Accuracy of the model on the training partition (percentage of correctly classified instances).", new PercentValue()));
|
|---|
| 145 | Add(new Result(TestAccuracyResultName, "Accuracy of the model on the test partition (percentage of correctly classified instances).", new PercentValue()));
|
|---|
| [6913] | 146 | Add(new Result(TrainingNormalizedGiniCoefficientResultName, "Normalized Gini coefficient of the model on the training partition.", new DoubleValue()));
|
|---|
| 147 | Add(new Result(TestNormalizedGiniCoefficientResultName, "Normalized Gini coefficient of the model on the test partition.", new DoubleValue()));
|
|---|
| [11622] | 148 | AddQualityMeasuresResultCollection();
|
|---|
| [6589] | 149 | }
|
|---|
| 150 |
|
|---|
| [7011] | 151 | [StorableHook(HookType.AfterDeserialization)]
|
|---|
| 152 | private void AfterDeserialization() {
|
|---|
| 153 | if (!this.ContainsKey(TrainingNormalizedGiniCoefficientResultName))
|
|---|
| 154 | Add(new Result(TrainingNormalizedGiniCoefficientResultName, "Normalized Gini coefficient of the model on the training partition.", new DoubleValue()));
|
|---|
| 155 | if (!this.ContainsKey(TestNormalizedGiniCoefficientResultName))
|
|---|
| 156 | Add(new Result(TestNormalizedGiniCoefficientResultName, "Normalized Gini coefficient of the model on the test partition.", new DoubleValue()));
|
|---|
| [11622] | 157 | if (!this.ContainsKey(QualityMeasuresResultName))
|
|---|
| 158 | AddQualityMeasuresResultCollection();
|
|---|
| [7011] | 159 | }
|
|---|
| 160 |
|
|---|
| [11622] | 161 | protected void AddQualityMeasuresResultCollection() {
|
|---|
| 162 | ResultCollection qualityMeasuresResult = new ResultCollection();
|
|---|
| 163 | qualityMeasuresResult.Add(new Result(TrainingTruePositiveRateResultName, "Sensitivity/True positive rate of the model on the training partition\n(TP/(TP+FN)).", new PercentValue()));
|
|---|
| 164 | qualityMeasuresResult.Add(new Result(TrainingTrueNegativeRateResultName, "Specificity/True negative rate of the model on the training partition\n(TN/(FP+TN)).", new PercentValue()));
|
|---|
| 165 | qualityMeasuresResult.Add(new Result(TrainingPositivePredictiveValueResultName, "Precision/Positive predictive value of the model on the training partition\n(TP/(TP+FP)).", new PercentValue()));
|
|---|
| 166 | qualityMeasuresResult.Add(new Result(TrainingNegativePredictiveValueResultName, "Negative predictive value of the model on the training partition\n(TN/(TN+FN)).", new PercentValue()));
|
|---|
| 167 | qualityMeasuresResult.Add(new Result(TrainingFalsePositiveRateResultName, "The false positive rate is the complement of the true negative rate of the model on the training partition.", new PercentValue()));
|
|---|
| 168 | qualityMeasuresResult.Add(new Result(TrainingFalseDiscoveryRateResultName, "The false discovery rate is the complement of the positive predictive value of the model on the training partition.", new PercentValue()));
|
|---|
| 169 | qualityMeasuresResult.Add(new Result(TestTruePositiveRateResultName, "Sensitivity/True positive rate of the model on the test partition\n(TP/(TP+FN)).", new PercentValue()));
|
|---|
| 170 | qualityMeasuresResult.Add(new Result(TestTrueNegativeRateResultName, "Specificity/True negative rate of the model on the test partition\n(TN/(FP+TN)).", new PercentValue()));
|
|---|
| 171 | qualityMeasuresResult.Add(new Result(TestPositivePredictiveValueResultName, "Precision/Positive predictive value of the model on the test partition\n(TP/(TP+FP)).", new PercentValue()));
|
|---|
| 172 | qualityMeasuresResult.Add(new Result(TestNegativePredictiveValueResultName, "Negative predictive value of the model on the test partition\n(TN/(TN+FN)).", new PercentValue()));
|
|---|
| 173 | qualityMeasuresResult.Add(new Result(TestFalsePositiveRateResultName, "The false positive rate is the complement of the true negative rate of the model on the test partition.", new PercentValue()));
|
|---|
| 174 | qualityMeasuresResult.Add(new Result(TestFalseDiscoveryRateResultName, "The false discovery rate is the complement of the positive predictive value of the model on the test partition.", new PercentValue()));
|
|---|
| 175 | Add(new Result(QualityMeasuresResultName, "Classification quality measures.\nIn Multiclass Classification all misclassifications of the negative class will be treated as true negatives.", qualityMeasuresResult));
|
|---|
| 176 | }
|
|---|
| 177 |
|
|---|
| [8723] | 178 | protected void CalculateClassificationResults() {
|
|---|
| [6589] | 179 | double[] estimatedTrainingClassValues = EstimatedTrainingClassValues.ToArray(); // cache values
|
|---|
| [8139] | 180 | double[] originalTrainingClassValues = ProblemData.Dataset.GetDoubleValues(ProblemData.TargetVariable, ProblemData.TrainingIndices).ToArray();
|
|---|
| [11622] | 181 |
|
|---|
| [6589] | 182 | double[] estimatedTestClassValues = EstimatedTestClassValues.ToArray(); // cache values
|
|---|
| [8139] | 183 | double[] originalTestClassValues = ProblemData.Dataset.GetDoubleValues(ProblemData.TargetVariable, ProblemData.TestIndices).ToArray();
|
|---|
| [6589] | 184 |
|
|---|
| [11622] | 185 | var positiveClassName = ProblemData.PositiveClassName;
|
|---|
| 186 | double positiveClassValue = ProblemData.GetClassValue(positiveClassName);
|
|---|
| 187 | QualityCalculator trainingQualityCalculator = new QualityCalculator(positiveClassValue);
|
|---|
| 188 | QualityCalculator testQualityCalculator = new QualityCalculator(positiveClassValue);
|
|---|
| 189 |
|
|---|
| [6589] | 190 | OnlineCalculatorError errorState;
|
|---|
| [6961] | 191 | double trainingAccuracy = OnlineAccuracyCalculator.Calculate(originalTrainingClassValues, estimatedTrainingClassValues, out errorState);
|
|---|
| [6589] | 192 | if (errorState != OnlineCalculatorError.None) trainingAccuracy = double.NaN;
|
|---|
| [6961] | 193 | double testAccuracy = OnlineAccuracyCalculator.Calculate(originalTestClassValues, estimatedTestClassValues, out errorState);
|
|---|
| [6589] | 194 | if (errorState != OnlineCalculatorError.None) testAccuracy = double.NaN;
|
|---|
| 195 |
|
|---|
| 196 | TrainingAccuracy = trainingAccuracy;
|
|---|
| 197 | TestAccuracy = testAccuracy;
|
|---|
| [6913] | 198 |
|
|---|
| 199 | double trainingNormalizedGini = NormalizedGiniCalculator.Calculate(originalTrainingClassValues, estimatedTrainingClassValues, out errorState);
|
|---|
| 200 | if (errorState != OnlineCalculatorError.None) trainingNormalizedGini = double.NaN;
|
|---|
| 201 | double testNormalizedGini = NormalizedGiniCalculator.Calculate(originalTestClassValues, estimatedTestClassValues, out errorState);
|
|---|
| 202 | if (errorState != OnlineCalculatorError.None) testNormalizedGini = double.NaN;
|
|---|
| 203 |
|
|---|
| 204 | TrainingNormalizedGiniCoefficient = trainingNormalizedGini;
|
|---|
| 205 | TestNormalizedGiniCoefficient = testNormalizedGini;
|
|---|
| [11622] | 206 |
|
|---|
| 207 | //quality measures training partition
|
|---|
| 208 | trainingQualityCalculator.Calculate(originalTrainingClassValues, estimatedTrainingClassValues, out errorState);
|
|---|
| 209 | if (errorState != OnlineCalculatorError.None) {
|
|---|
| 210 | TrainingTruePositiveRate = double.NaN;
|
|---|
| 211 | TrainingTrueNegativeRate = double.NaN;
|
|---|
| 212 | TrainingPositivePredictiveValue = double.NaN;
|
|---|
| 213 | TrainingNegativePredictiveValue = double.NaN;
|
|---|
| 214 | TrainingFalsePositiveRate = double.NaN;
|
|---|
| 215 | TrainingFalseDiscoveryRate = double.NaN;
|
|---|
| 216 | } else {
|
|---|
| 217 | TrainingTruePositiveRate = trainingQualityCalculator.TruePositiveRate;
|
|---|
| 218 | TrainingTrueNegativeRate = trainingQualityCalculator.TrueNegativeRate;
|
|---|
| 219 | TrainingPositivePredictiveValue = trainingQualityCalculator.PositivePredictiveValue;
|
|---|
| 220 | TrainingNegativePredictiveValue = trainingQualityCalculator.NegativePredictiveValue;
|
|---|
| 221 | TrainingFalsePositiveRate = trainingQualityCalculator.FalsePositiveRate;
|
|---|
| 222 | TrainingFalseDiscoveryRate = trainingQualityCalculator.FalseDiscoveryRate;
|
|---|
| 223 | }
|
|---|
| 224 | //quality measures test partition
|
|---|
| 225 | testQualityCalculator.Calculate(originalTestClassValues, estimatedTestClassValues, out errorState);
|
|---|
| 226 | if (errorState != OnlineCalculatorError.None) {
|
|---|
| 227 | TestTruePositiveRate = double.NaN;
|
|---|
| 228 | TestTrueNegativeRate = double.NaN;
|
|---|
| 229 | TestPositivePredictiveValue = double.NaN;
|
|---|
| 230 | TestNegativePredictiveValue = double.NaN;
|
|---|
| 231 | TestFalsePositiveRate = double.NaN;
|
|---|
| 232 | TestFalseDiscoveryRate = double.NaN;
|
|---|
| 233 | } else {
|
|---|
| 234 | TestTruePositiveRate = testQualityCalculator.TruePositiveRate;
|
|---|
| 235 | TestTrueNegativeRate = testQualityCalculator.TrueNegativeRate;
|
|---|
| 236 | TestPositivePredictiveValue = testQualityCalculator.PositivePredictiveValue;
|
|---|
| 237 | TestNegativePredictiveValue = testQualityCalculator.NegativePredictiveValue;
|
|---|
| 238 | TestFalsePositiveRate = testQualityCalculator.FalsePositiveRate;
|
|---|
| 239 | TestFalseDiscoveryRate = testQualityCalculator.FalseDiscoveryRate;
|
|---|
| 240 | }
|
|---|
| [6589] | 241 | }
|
|---|
| 242 |
|
|---|
| 243 | public abstract IEnumerable<double> EstimatedClassValues { get; }
|
|---|
| 244 | public abstract IEnumerable<double> EstimatedTrainingClassValues { get; }
|
|---|
| 245 | public abstract IEnumerable<double> EstimatedTestClassValues { get; }
|
|---|
| 246 |
|
|---|
| 247 | public abstract IEnumerable<double> GetEstimatedClassValues(IEnumerable<int> rows);
|
|---|
| [8723] | 248 |
|
|---|
| 249 | protected override void RecalculateResults() {
|
|---|
| 250 | CalculateClassificationResults();
|
|---|
| 251 | }
|
|---|
| [6589] | 252 | }
|
|---|
| 253 | }
|
|---|
