[5275] | 1 | #region License Information
|
---|
| 2 | /* HeuristicLab
|
---|
| 3 | * Copyright (C) 2002-2010 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.Collections.Generic;
|
---|
| 23 | using System.Linq;
|
---|
| 24 | using HeuristicLab.Analysis;
|
---|
| 25 | using HeuristicLab.Common;
|
---|
| 26 | using HeuristicLab.Core;
|
---|
| 27 | using HeuristicLab.Data;
|
---|
| 28 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
|
---|
| 29 | using HeuristicLab.Operators;
|
---|
| 30 | using HeuristicLab.Optimization;
|
---|
| 31 | using HeuristicLab.Parameters;
|
---|
| 32 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
| 33 | using HeuristicLab.Problems.DataAnalysis.Evaluators;
|
---|
| 34 | using HeuristicLab.Problems.DataAnalysis.Symbolic;
|
---|
| 35 | using System;
|
---|
| 36 |
|
---|
| 37 | namespace HeuristicLab.Problems.DataAnalysis.Regression.Symbolic.Analyzers {
|
---|
| 38 | [Item("SymbolicRegressionOverfittingAnalyzer", "Calculates and tracks correlation of training and validation fitness of symbolic regression models.")]
|
---|
| 39 | [StorableClass]
|
---|
| 40 | public sealed class SymbolicRegressionOverfittingAnalyzer : SymbolicRegressionValidationAnalyzer, ISymbolicRegressionAnalyzer {
|
---|
| 41 | private const string MaximizationParameterName = "Maximization";
|
---|
| 42 | private const string QualityParameterName = "Quality";
|
---|
| 43 | private const string TrainingValidationCorrelationParameterName = "TrainingValidationCorrelation";
|
---|
| 44 | private const string TrainingValidationCorrelationTableParameterName = "TrainingValidationCorrelationTable";
|
---|
| 45 | private const string LowerCorrelationThresholdParameterName = "LowerCorrelationThreshold";
|
---|
| 46 | private const string UpperCorrelationThresholdParameterName = "UpperCorrelationThreshold";
|
---|
| 47 | private const string OverfittingParameterName = "IsOverfitting";
|
---|
| 48 | private const string ResultsParameterName = "Results";
|
---|
| 49 |
|
---|
| 50 | #region parameter properties
|
---|
| 51 | public ScopeTreeLookupParameter<DoubleValue> QualityParameter {
|
---|
| 52 | get { return (ScopeTreeLookupParameter<DoubleValue>)Parameters[QualityParameterName]; }
|
---|
| 53 | }
|
---|
| 54 | public ILookupParameter<BoolValue> MaximizationParameter {
|
---|
| 55 | get { return (ILookupParameter<BoolValue>)Parameters[MaximizationParameterName]; }
|
---|
| 56 | }
|
---|
| 57 | public ILookupParameter<DoubleValue> TrainingValidationQualityCorrelationParameter {
|
---|
| 58 | get { return (ILookupParameter<DoubleValue>)Parameters[TrainingValidationCorrelationParameterName]; }
|
---|
| 59 | }
|
---|
| 60 | public ILookupParameter<DataTable> TrainingValidationQualityCorrelationTableParameter {
|
---|
| 61 | get { return (ILookupParameter<DataTable>)Parameters[TrainingValidationCorrelationTableParameterName]; }
|
---|
| 62 | }
|
---|
| 63 | public IValueLookupParameter<DoubleValue> LowerCorrelationThresholdParameter {
|
---|
| 64 | get { return (IValueLookupParameter<DoubleValue>)Parameters[LowerCorrelationThresholdParameterName]; }
|
---|
| 65 | }
|
---|
| 66 | public IValueLookupParameter<DoubleValue> UpperCorrelationThresholdParameter {
|
---|
| 67 | get { return (IValueLookupParameter<DoubleValue>)Parameters[UpperCorrelationThresholdParameterName]; }
|
---|
| 68 | }
|
---|
| 69 | public ILookupParameter<BoolValue> OverfittingParameter {
|
---|
| 70 | get { return (ILookupParameter<BoolValue>)Parameters[OverfittingParameterName]; }
|
---|
| 71 | }
|
---|
| 72 | public ILookupParameter<ResultCollection> ResultsParameter {
|
---|
| 73 | get { return (ILookupParameter<ResultCollection>)Parameters[ResultsParameterName]; }
|
---|
| 74 | }
|
---|
| 75 | #endregion
|
---|
| 76 | #region properties
|
---|
| 77 | public BoolValue Maximization {
|
---|
| 78 | get { return MaximizationParameter.ActualValue; }
|
---|
| 79 | }
|
---|
| 80 | #endregion
|
---|
| 81 |
|
---|
| 82 | [StorableConstructor]
|
---|
| 83 | private SymbolicRegressionOverfittingAnalyzer(bool deserializing) : base(deserializing) { }
|
---|
| 84 | private SymbolicRegressionOverfittingAnalyzer(SymbolicRegressionOverfittingAnalyzer original, Cloner cloner) : base(original, cloner) { }
|
---|
| 85 | public SymbolicRegressionOverfittingAnalyzer()
|
---|
| 86 | : base() {
|
---|
| 87 | Parameters.Add(new ScopeTreeLookupParameter<DoubleValue>(QualityParameterName, "Training fitness"));
|
---|
| 88 | Parameters.Add(new LookupParameter<BoolValue>(MaximizationParameterName, "The direction of optimization."));
|
---|
| 89 | Parameters.Add(new LookupParameter<DoubleValue>(TrainingValidationCorrelationParameterName, "Correlation of training and validation fitnesses"));
|
---|
| 90 | Parameters.Add(new LookupParameter<DataTable>(TrainingValidationCorrelationTableParameterName, "Data table of training and validation fitness correlation values over the whole run."));
|
---|
| 91 | Parameters.Add(new ValueLookupParameter<DoubleValue>(LowerCorrelationThresholdParameterName, "Lower threshold for correlation value that marks the boundary from non-overfitting to overfitting.", new DoubleValue(0.65)));
|
---|
| 92 | Parameters.Add(new ValueLookupParameter<DoubleValue>(UpperCorrelationThresholdParameterName, "Upper threshold for correlation value that marks the boundary from overfitting to non-overfitting.", new DoubleValue(0.75)));
|
---|
| 93 | Parameters.Add(new LookupParameter<BoolValue>(OverfittingParameterName, "Boolean indicator for overfitting."));
|
---|
| 94 | Parameters.Add(new LookupParameter<ResultCollection>(ResultsParameterName, "The results collection."));
|
---|
| 95 | }
|
---|
| 96 |
|
---|
| 97 | [StorableHook(HookType.AfterDeserialization)]
|
---|
| 98 | private void AfterDeserialization() {
|
---|
| 99 | }
|
---|
| 100 |
|
---|
| 101 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
| 102 | return new SymbolicRegressionOverfittingAnalyzer(this, cloner);
|
---|
| 103 | }
|
---|
| 104 |
|
---|
| 105 | protected override void Analyze(SymbolicExpressionTree[] trees, double[] validationQuality) {
|
---|
| 106 | double[] trainingQuality = QualityParameter.ActualValue.Select(x => x.Value).ToArray();
|
---|
| 107 |
|
---|
| 108 | double r = alglib.spearmancorr2(trainingQuality, validationQuality);
|
---|
| 109 |
|
---|
| 110 | TrainingValidationQualityCorrelationParameter.ActualValue = new DoubleValue(r);
|
---|
| 111 |
|
---|
| 112 | if (TrainingValidationQualityCorrelationTableParameter.ActualValue == null) {
|
---|
| 113 | var dataTable = new DataTable("Training and validation fitness correlation table", "Data table of training and validation fitness correlation values over the whole run.");
|
---|
| 114 | dataTable.Rows.Add(new DataRow("Training and validation fitness correlation", "Training and validation fitness correlation values"));
|
---|
| 115 | TrainingValidationQualityCorrelationTableParameter.ActualValue = dataTable;
|
---|
| 116 | ResultsParameter.ActualValue.Add(new Result(TrainingValidationCorrelationTableParameterName, dataTable));
|
---|
| 117 | }
|
---|
| 118 |
|
---|
| 119 | TrainingValidationQualityCorrelationTableParameter.ActualValue.Rows["Training and validation fitness correlation"].Values.Add(r);
|
---|
| 120 |
|
---|
| 121 | double correlationThreshold;
|
---|
| 122 | if (OverfittingParameter.ActualValue != null && OverfittingParameter.ActualValue.Value) {
|
---|
| 123 | // if is already overfitting => have to reach the upper threshold to switch back to non-overfitting state
|
---|
| 124 | correlationThreshold = UpperCorrelationThresholdParameter.ActualValue.Value;
|
---|
| 125 | } else {
|
---|
| 126 | // if currently in non-overfitting state => have to reach to lower threshold to switch to overfitting state
|
---|
| 127 | correlationThreshold = LowerCorrelationThresholdParameter.ActualValue.Value;
|
---|
| 128 | }
|
---|
| 129 | bool overfitting = r < correlationThreshold;
|
---|
| 130 |
|
---|
| 131 | OverfittingParameter.ActualValue = new BoolValue(overfitting);
|
---|
| 132 | }
|
---|
| 133 | }
|
---|
| 134 | }
|
---|