[14630] | 1 | #region License Information
|
---|
| 2 | /* HeuristicLab
|
---|
[16565] | 3 | * Copyright (C) 2002-2019 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
[14630] | 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 System.Linq;
|
---|
| 25 | using HeuristicLab.Common;
|
---|
| 26 | using HeuristicLab.Core;
|
---|
| 27 | using HeuristicLab.Problems.DataAnalysis;
|
---|
| 28 | using HeuristicLab.Random;
|
---|
| 29 |
|
---|
| 30 | namespace HeuristicLab.Problems.Instances.DataAnalysis {
|
---|
| 31 | public sealed class LinearVariableNetwork : VariableNetwork {
|
---|
| 32 | private int numberOfFeatures;
|
---|
| 33 | private double noiseRatio;
|
---|
| 34 |
|
---|
| 35 | public override string Name { get { return string.Format("LinearVariableNetwork-{0:0%} ({1} dim)", noiseRatio, numberOfFeatures); } }
|
---|
| 36 |
|
---|
| 37 | public LinearVariableNetwork(int numberOfFeatures, double noiseRatio,
|
---|
| 38 | IRandom rand)
|
---|
| 39 | : base(250, 250, numberOfFeatures, noiseRatio, rand) {
|
---|
| 40 | this.noiseRatio = noiseRatio;
|
---|
| 41 | this.numberOfFeatures = numberOfFeatures;
|
---|
| 42 | }
|
---|
| 43 |
|
---|
| 44 | protected override IEnumerable<double> GenerateRandomFunction(IRandom rand, List<List<double>> xs, out string[] selectedVarNames, out double[] relevance) {
|
---|
| 45 | int nl = SampleNumberOfVariables(rand, numberOfFeatures);
|
---|
| 46 |
|
---|
| 47 | var selectedIdx = Enumerable.Range(0, xs.Count).Shuffle(rand)
|
---|
| 48 | .Take(nl).ToArray();
|
---|
| 49 |
|
---|
| 50 | var selectedVars = selectedIdx.Select(i => xs[i]).ToArray();
|
---|
| 51 | selectedVarNames = selectedIdx.Select(i => VariableNames[i]).ToArray();
|
---|
| 52 | return SampleLinearFunction(rand, selectedVars, out relevance);
|
---|
| 53 | }
|
---|
| 54 |
|
---|
| 55 | private IEnumerable<double> SampleLinearFunction(IRandom rand, List<double>[] xs, out double[] relevance) {
|
---|
| 56 | int nl = xs.Length;
|
---|
| 57 | int nRows = xs.First().Count;
|
---|
| 58 |
|
---|
| 59 | // sample standardized coefficients iid ~ N(0, 1)
|
---|
| 60 | var c = Enumerable.Range(0, nRows).Select(_ => NormalDistributedRandom.NextDouble(rand, 0, 1)).ToArray();
|
---|
| 61 |
|
---|
| 62 | // calculate scaled coefficients (variables with large variance should have smaller coefficients)
|
---|
| 63 | var scaledC = Enumerable.Range(0, nl)
|
---|
| 64 | .Select(i => c[i] / xs[i].StandardDeviationPop())
|
---|
| 65 | .ToArray();
|
---|
| 66 |
|
---|
| 67 | var y = EvaluteLinearModel(xs, scaledC);
|
---|
| 68 |
|
---|
| 69 | relevance = CalculateRelevance(y, xs, scaledC);
|
---|
| 70 |
|
---|
| 71 | return y;
|
---|
| 72 | }
|
---|
| 73 |
|
---|
| 74 | private double[] EvaluteLinearModel(List<double>[] xs, double[] c) {
|
---|
| 75 | int nRows = xs.First().Count;
|
---|
| 76 | var y = new double[nRows];
|
---|
| 77 | for(int row = 0; row < nRows; row++) {
|
---|
| 78 | y[row] = xs.Select(xi => xi[row]).Zip(c, (xij, cj) => xij * cj).Sum();
|
---|
| 79 | y[row] /= c.Length;
|
---|
| 80 | }
|
---|
| 81 | return y;
|
---|
| 82 | }
|
---|
| 83 |
|
---|
| 84 | // calculate variable relevance based on removal of variables
|
---|
| 85 | // 1) to remove a variable we set it's coefficient to zero
|
---|
| 86 | // 2) calculate MSE of the original target values (y) to the updated targes y' (after variable removal)
|
---|
| 87 | // 3) relevance is larger if MSE(y,y') is large
|
---|
| 88 | // 4) scale impacts so that the most important variable has impact = 1
|
---|
| 89 | private double[] CalculateRelevance(double[] y, List<double>[] xs, double[] l) {
|
---|
| 90 | var changedL = new double[l.Length];
|
---|
| 91 | var relevance = new double[l.Length];
|
---|
| 92 | for(int i = 0; i < l.Length; i++) {
|
---|
| 93 | Array.Copy(l, changedL, changedL.Length);
|
---|
| 94 | changedL[i] = 0.0;
|
---|
| 95 |
|
---|
| 96 | var yChanged = EvaluteLinearModel(xs, changedL);
|
---|
| 97 |
|
---|
| 98 | OnlineCalculatorError error;
|
---|
| 99 | var mse = OnlineMeanSquaredErrorCalculator.Calculate(y, yChanged, out error);
|
---|
| 100 | if(error != OnlineCalculatorError.None) mse = double.MaxValue;
|
---|
| 101 | relevance[i] = mse;
|
---|
| 102 | }
|
---|
| 103 | // scale so that max relevance is 1.0
|
---|
| 104 | var maxRel = relevance.Max();
|
---|
| 105 | for(int i = 0; i < relevance.Length; i++) relevance[i] /= maxRel;
|
---|
| 106 | return relevance;
|
---|
| 107 | }
|
---|
| 108 | }
|
---|
| 109 | }
|
---|