[8798] | 1 | #region License Information
|
---|
| 2 | /* HeuristicLab
|
---|
[11171] | 3 | * Copyright (C) 2002-2014 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
[8798] | 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 |
|
---|
[9452] | 22 | using System;
|
---|
[8798] | 23 | using System.Linq;
|
---|
| 24 | using HeuristicLab.Common;
|
---|
| 25 | using HeuristicLab.Core;
|
---|
| 26 | using HeuristicLab.Parameters;
|
---|
| 27 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
[9452] | 28 | using HeuristicLab.Problems.Instances;
|
---|
[8798] | 29 |
|
---|
| 30 | namespace HeuristicLab.Problems.DataAnalysis.Symbolic.TimeSeriesPrognosis {
|
---|
| 31 | [Item("Symbolic Time-Series Prognosis Problem (single objective)", "Represents a single objective symbolic time-series prognosis problem.")]
|
---|
| 32 | [StorableClass]
|
---|
| 33 | [Creatable("Problems")]
|
---|
| 34 | public class SymbolicTimeSeriesPrognosisSingleObjectiveProblem : SymbolicDataAnalysisSingleObjectiveProblem<ITimeSeriesPrognosisProblemData, ISymbolicTimeSeriesPrognosisSingleObjectiveEvaluator, ISymbolicDataAnalysisSolutionCreator>, ITimeSeriesPrognosisProblem {
|
---|
| 35 | private const double PunishmentFactor = 10;
|
---|
| 36 | private const int InitialMaximumTreeDepth = 8;
|
---|
| 37 | private const int InitialMaximumTreeLength = 25;
|
---|
| 38 | private const string EstimationLimitsParameterName = "EstimationLimits";
|
---|
| 39 | private const string EstimationLimitsParameterDescription = "The limits for the estimated value that can be returned by the symbolic regression model.";
|
---|
| 40 |
|
---|
| 41 | #region parameter properties
|
---|
| 42 | public IFixedValueParameter<DoubleLimit> EstimationLimitsParameter {
|
---|
| 43 | get { return (IFixedValueParameter<DoubleLimit>)Parameters[EstimationLimitsParameterName]; }
|
---|
| 44 | }
|
---|
| 45 | #endregion
|
---|
| 46 | #region properties
|
---|
| 47 | public DoubleLimit EstimationLimits {
|
---|
| 48 | get { return EstimationLimitsParameter.Value; }
|
---|
| 49 | }
|
---|
| 50 | #endregion
|
---|
| 51 | [StorableConstructor]
|
---|
| 52 | protected SymbolicTimeSeriesPrognosisSingleObjectiveProblem(bool deserializing) : base(deserializing) { }
|
---|
| 53 | protected SymbolicTimeSeriesPrognosisSingleObjectiveProblem(SymbolicTimeSeriesPrognosisSingleObjectiveProblem original, Cloner cloner) : base(original, cloner) { }
|
---|
| 54 | public override IDeepCloneable Clone(Cloner cloner) { return new SymbolicTimeSeriesPrognosisSingleObjectiveProblem(this, cloner); }
|
---|
| 55 |
|
---|
| 56 | public SymbolicTimeSeriesPrognosisSingleObjectiveProblem()
|
---|
| 57 | : base(new TimeSeriesPrognosisProblemData(), new SymbolicTimeSeriesPrognosisSingleObjectiveMeanSquaredErrorEvaluator(), new SymbolicDataAnalysisExpressionTreeCreator()) {
|
---|
| 58 | Parameters.Add(new FixedValueParameter<DoubleLimit>(EstimationLimitsParameterName, EstimationLimitsParameterDescription));
|
---|
| 59 | EstimationLimitsParameter.Hidden = true;
|
---|
| 60 |
|
---|
| 61 | Maximization.Value = false;
|
---|
| 62 | MaximumSymbolicExpressionTreeDepth.Value = InitialMaximumTreeDepth;
|
---|
| 63 | MaximumSymbolicExpressionTreeLength.Value = InitialMaximumTreeLength;
|
---|
| 64 |
|
---|
| 65 | var interpeter = new SymbolicTimeSeriesPrognosisExpressionTreeInterpreter();
|
---|
| 66 | interpeter.TargetVariable = ProblemData.TargetVariable;
|
---|
| 67 | SymbolicExpressionTreeInterpreter = interpeter;
|
---|
| 68 |
|
---|
| 69 | SymbolicExpressionTreeGrammarParameter.ValueChanged += (o, e) => ConfigureGrammarSymbols();
|
---|
| 70 | ConfigureGrammarSymbols();
|
---|
| 71 |
|
---|
| 72 | InitializeOperators();
|
---|
| 73 | UpdateEstimationLimits();
|
---|
| 74 | }
|
---|
| 75 |
|
---|
| 76 | private void ConfigureGrammarSymbols() {
|
---|
| 77 | var grammar = SymbolicExpressionTreeGrammar as TypeCoherentExpressionGrammar;
|
---|
| 78 | if (grammar != null) grammar.ConfigureAsDefaultTimeSeriesPrognosisGrammar();
|
---|
| 79 | UpdateGrammar();
|
---|
| 80 | }
|
---|
| 81 | protected override void UpdateGrammar() {
|
---|
| 82 | base.UpdateGrammar();
|
---|
| 83 | foreach (var autoregressiveSymbol in SymbolicExpressionTreeGrammar.Symbols.OfType<AutoregressiveTargetVariable>()) {
|
---|
| 84 | if (!autoregressiveSymbol.Fixed) autoregressiveSymbol.VariableNames = ProblemData.TargetVariable.ToEnumerable();
|
---|
| 85 | }
|
---|
| 86 | }
|
---|
| 87 |
|
---|
| 88 | private void InitializeOperators() {
|
---|
| 89 | Operators.Add(new SymbolicTimeSeriesPrognosisSingleObjectiveTrainingBestSolutionAnalyzer());
|
---|
| 90 | Operators.Add(new SymbolicTimeSeriesPrognosisSingleObjectiveValidationBestSolutionAnalyzer());
|
---|
| 91 | Operators.Add(new SymbolicTimeSeriesPrognosisSingleObjectiveOverfittingAnalyzer());
|
---|
| 92 | ParameterizeOperators();
|
---|
| 93 | }
|
---|
| 94 |
|
---|
| 95 | private void UpdateEstimationLimits() {
|
---|
| 96 | if (ProblemData.TrainingIndices.Any()) {
|
---|
| 97 | var targetValues = ProblemData.Dataset.GetDoubleValues(ProblemData.TargetVariable, ProblemData.TrainingIndices).ToList();
|
---|
| 98 | var mean = targetValues.Average();
|
---|
| 99 | var range = targetValues.Max() - targetValues.Min();
|
---|
| 100 | EstimationLimits.Upper = mean + PunishmentFactor * range;
|
---|
| 101 | EstimationLimits.Lower = mean - PunishmentFactor * range;
|
---|
| 102 | } else {
|
---|
| 103 | EstimationLimits.Upper = double.MaxValue;
|
---|
| 104 | EstimationLimits.Lower = double.MinValue;
|
---|
| 105 | }
|
---|
| 106 | }
|
---|
| 107 |
|
---|
| 108 | protected override void OnProblemDataChanged() {
|
---|
| 109 | base.OnProblemDataChanged();
|
---|
| 110 | var interpreter = SymbolicExpressionTreeInterpreter as ISymbolicTimeSeriesPrognosisExpressionTreeInterpreter;
|
---|
| 111 | if (interpreter != null) {
|
---|
| 112 | interpreter.TargetVariable = ProblemData.TargetVariable;
|
---|
| 113 | }
|
---|
| 114 | UpdateEstimationLimits();
|
---|
| 115 |
|
---|
| 116 | }
|
---|
| 117 |
|
---|
| 118 | protected override void ParameterizeOperators() {
|
---|
| 119 | base.ParameterizeOperators();
|
---|
| 120 | if (Parameters.ContainsKey(EstimationLimitsParameterName)) {
|
---|
| 121 | var operators = Parameters.OfType<IValueParameter>().Select(p => p.Value).OfType<IOperator>().Union(Operators);
|
---|
| 122 | foreach (var op in operators.OfType<ISymbolicDataAnalysisBoundedOperator>()) {
|
---|
| 123 | op.EstimationLimitsParameter.ActualName = EstimationLimitsParameter.Name;
|
---|
| 124 | }
|
---|
| 125 | foreach (var op in operators.OfType<SymbolicTimeSeriesPrognosisSingleObjectiveTrainingBestSolutionAnalyzer>()) {
|
---|
| 126 | op.MaximizationParameter.ActualName = MaximizationParameter.Name;
|
---|
| 127 | op.ProblemDataParameter.ActualName = ProblemDataParameter.Name;
|
---|
| 128 | op.QualityParameter.ActualName = Evaluator.QualityParameter.ActualName;
|
---|
| 129 | op.SymbolicDataAnalysisTreeInterpreterParameter.ActualName = SymbolicExpressionTreeInterpreterParameter.Name;
|
---|
| 130 | op.SymbolicExpressionTreeParameter.ActualName = SolutionCreator.SymbolicExpressionTreeParameter.ActualName;
|
---|
| 131 | }
|
---|
| 132 | }
|
---|
| 133 | }
|
---|
[9452] | 134 |
|
---|
| 135 | public override void Load(ITimeSeriesPrognosisProblemData data) {
|
---|
| 136 | base.Load(data);
|
---|
| 137 | this.ProblemData.TrainingPartition.Start =
|
---|
| 138 | Math.Min(10, this.ProblemData.TrainingPartition.End);
|
---|
| 139 | }
|
---|
[8798] | 140 | }
|
---|
| 141 | }
|
---|