[5624] | 1 | #region License Information
|
---|
| 2 | /* HeuristicLab
|
---|
[7259] | 3 | * Copyright (C) 2002-2012 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
[5624] | 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 HeuristicLab.Common;
|
---|
| 24 | using HeuristicLab.Core;
|
---|
| 25 | using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
|
---|
| 26 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
| 27 |
|
---|
| 28 | namespace HeuristicLab.Problems.DataAnalysis.Symbolic.Classification {
|
---|
| 29 | /// <summary>
|
---|
| 30 | /// Represents a symbolic classification model
|
---|
| 31 | /// </summary>
|
---|
| 32 | [StorableClass]
|
---|
| 33 | [Item(Name = "SymbolicClassificationModel", Description = "Represents a symbolic classification model.")]
|
---|
[8660] | 34 | public abstract class
|
---|
| 35 | SymbolicClassificationModel : SymbolicDataAnalysisModel, ISymbolicClassificationModel {
|
---|
| 36 | [Storable]
|
---|
| 37 | private double lowerEstimationLimit;
|
---|
| 38 | public double LowerEstimationLimit { get { return lowerEstimationLimit; } }
|
---|
| 39 | [Storable]
|
---|
| 40 | private double upperEstimationLimit;
|
---|
| 41 | public double UpperEstimationLimit { get { return upperEstimationLimit; } }
|
---|
| 42 |
|
---|
[5624] | 43 | [StorableConstructor]
|
---|
| 44 | protected SymbolicClassificationModel(bool deserializing) : base(deserializing) { }
|
---|
| 45 | protected SymbolicClassificationModel(SymbolicClassificationModel original, Cloner cloner)
|
---|
| 46 | : base(original, cloner) {
|
---|
[8660] | 47 | lowerEstimationLimit = original.lowerEstimationLimit;
|
---|
| 48 | upperEstimationLimit = original.upperEstimationLimit;
|
---|
[5624] | 49 | }
|
---|
[8660] | 50 | protected SymbolicClassificationModel(ISymbolicExpressionTree tree, ISymbolicDataAnalysisExpressionTreeInterpreter interpreter, double lowerEstimationLimit = double.MinValue, double upperEstimationLimit = double.MaxValue)
|
---|
[5624] | 51 | : base(tree, interpreter) {
|
---|
[8660] | 52 | this.lowerEstimationLimit = lowerEstimationLimit;
|
---|
| 53 | this.upperEstimationLimit = upperEstimationLimit;
|
---|
[5624] | 54 | }
|
---|
| 55 |
|
---|
[8660] | 56 | public abstract IEnumerable<double> GetEstimatedClassValues(Dataset dataset, IEnumerable<int> rows);
|
---|
| 57 | public abstract void RecalculateModelParameters(IClassificationProblemData problemData, IEnumerable<int> rows);
|
---|
| 58 |
|
---|
| 59 | public abstract ISymbolicClassificationSolution CreateClassificationSolution(IClassificationProblemData problemData);
|
---|
| 60 |
|
---|
| 61 | IClassificationSolution IClassificationModel.CreateClassificationSolution(IClassificationProblemData problemData) {
|
---|
| 62 | return CreateClassificationSolution(problemData);
|
---|
[5624] | 63 | }
|
---|
| 64 |
|
---|
[8660] | 65 | #region scaling
|
---|
| 66 | public static void Scale(ISymbolicClassificationModel model, IClassificationProblemData problemData) {
|
---|
| 67 | var dataset = problemData.Dataset;
|
---|
| 68 | var targetVariable = problemData.TargetVariable;
|
---|
| 69 | var rows = problemData.TrainingIndices;
|
---|
| 70 | var estimatedValues = model.Interpreter.GetSymbolicExpressionTreeValues(model.SymbolicExpressionTree, dataset, rows);
|
---|
| 71 | var targetValues = dataset.GetDoubleValues(targetVariable, rows);
|
---|
| 72 | double alpha;
|
---|
| 73 | double beta;
|
---|
| 74 | OnlineCalculatorError errorState;
|
---|
| 75 | OnlineLinearScalingParameterCalculator.Calculate(estimatedValues, targetValues, out alpha, out beta, out errorState);
|
---|
| 76 | if (errorState != OnlineCalculatorError.None) return;
|
---|
| 77 |
|
---|
| 78 | ConstantTreeNode alphaTreeNode = null;
|
---|
| 79 | ConstantTreeNode betaTreeNode = null;
|
---|
| 80 | // check if model has been scaled previously by analyzing the structure of the tree
|
---|
| 81 | var startNode = model.SymbolicExpressionTree.Root.GetSubtree(0);
|
---|
| 82 | if (startNode.GetSubtree(0).Symbol is Addition) {
|
---|
| 83 | var addNode = startNode.GetSubtree(0);
|
---|
| 84 | if (addNode.SubtreeCount == 2 && addNode.GetSubtree(0).Symbol is Multiplication && addNode.GetSubtree(1).Symbol is Constant) {
|
---|
| 85 | alphaTreeNode = addNode.GetSubtree(1) as ConstantTreeNode;
|
---|
| 86 | var mulNode = addNode.GetSubtree(0);
|
---|
| 87 | if (mulNode.SubtreeCount == 2 && mulNode.GetSubtree(1).Symbol is Constant) {
|
---|
| 88 | betaTreeNode = mulNode.GetSubtree(1) as ConstantTreeNode;
|
---|
| 89 | }
|
---|
| 90 | }
|
---|
| 91 | }
|
---|
| 92 | // if tree structure matches the structure necessary for linear scaling then reuse the existing tree nodes
|
---|
| 93 | if (alphaTreeNode != null && betaTreeNode != null) {
|
---|
| 94 | betaTreeNode.Value *= beta;
|
---|
| 95 | alphaTreeNode.Value *= beta;
|
---|
| 96 | alphaTreeNode.Value += alpha;
|
---|
| 97 | } else {
|
---|
| 98 | var mainBranch = startNode.GetSubtree(0);
|
---|
| 99 | startNode.RemoveSubtree(0);
|
---|
| 100 | var scaledMainBranch = MakeSum(MakeProduct(mainBranch, beta), alpha);
|
---|
| 101 | startNode.AddSubtree(scaledMainBranch);
|
---|
| 102 | }
|
---|
[5624] | 103 | }
|
---|
[6604] | 104 |
|
---|
[8660] | 105 | private static ISymbolicExpressionTreeNode MakeSum(ISymbolicExpressionTreeNode treeNode, double alpha) {
|
---|
| 106 | if (alpha.IsAlmost(0.0)) {
|
---|
| 107 | return treeNode;
|
---|
| 108 | } else {
|
---|
| 109 | var addition = new Addition();
|
---|
| 110 | var node = addition.CreateTreeNode();
|
---|
| 111 | var alphaConst = MakeConstant(alpha);
|
---|
| 112 | node.AddSubtree(treeNode);
|
---|
| 113 | node.AddSubtree(alphaConst);
|
---|
| 114 | return node;
|
---|
| 115 | }
|
---|
[6604] | 116 | }
|
---|
[8660] | 117 |
|
---|
| 118 | private static ISymbolicExpressionTreeNode MakeProduct(ISymbolicExpressionTreeNode treeNode, double beta) {
|
---|
| 119 | if (beta.IsAlmost(1.0)) {
|
---|
| 120 | return treeNode;
|
---|
| 121 | } else {
|
---|
| 122 | var multipliciation = new Multiplication();
|
---|
| 123 | var node = multipliciation.CreateTreeNode();
|
---|
| 124 | var betaConst = MakeConstant(beta);
|
---|
| 125 | node.AddSubtree(treeNode);
|
---|
| 126 | node.AddSubtree(betaConst);
|
---|
| 127 | return node;
|
---|
| 128 | }
|
---|
[6604] | 129 | }
|
---|
| 130 |
|
---|
[8660] | 131 | private static ISymbolicExpressionTreeNode MakeConstant(double c) {
|
---|
| 132 | var node = (ConstantTreeNode)(new Constant()).CreateTreeNode();
|
---|
| 133 | node.Value = c;
|
---|
| 134 | return node;
|
---|
| 135 | }
|
---|
| 136 | #endregion
|
---|
[5624] | 137 | }
|
---|
| 138 | }
|
---|