Free cookie consent management tool by TermsFeed Policy Generator

source: trunk/sources/HeuristicLab.Algorithms.DataAnalysis/3.4/GradientBoostedTrees/LossFunctions/SquaredErrorLoss.cs @ 12873

Last change on this file since 12873 was 12873, checked in by gkronber, 9 years ago

#2450 derived ILossFunction from IItem to allow execution on hive without privileged flag (made an "after deserialization"-hook necessary to convert the parameter type)

File size: 3.3 KB
Line 
1#region License Information
2/* HeuristicLab
3 * Copyright (C) 2002-2015 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
4 * and the BEACON Center for the Study of Evolution in Action.
5 *
6 * This file is part of HeuristicLab.
7 *
8 * HeuristicLab is free software: you can redistribute it and/or modify
9 * it under the terms of the GNU General Public License as published by
10 * the Free Software Foundation, either version 3 of the License, or
11 * (at your option) any later version.
12 *
13 * HeuristicLab is distributed in the hope that it will be useful,
14 * but WITHOUT ANY WARRANTY; without even the implied warranty of
15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
16 * GNU General Public License for more details.
17 *
18 * You should have received a copy of the GNU General Public License
19 * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
20 */
21#endregion
22
23using System;
24using System.Collections.Generic;
25using System.Linq;
26using HeuristicLab.Common;
27using HeuristicLab.Core;
28using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
29
30namespace HeuristicLab.Algorithms.DataAnalysis {
31  [StorableClass]
32  [Item("Squared error loss", "")]
33  public class SquaredErrorLoss : Item, ILossFunction {
34    public SquaredErrorLoss() { }
35
36    public double GetLoss(IEnumerable<double> target, IEnumerable<double> pred) {
37      var targetEnum = target.GetEnumerator();
38      var predEnum = pred.GetEnumerator();
39
40      double s = 0;
41      while (targetEnum.MoveNext() & predEnum.MoveNext()) {
42        double res = targetEnum.Current - predEnum.Current;
43        s += res * res; // (res)^2
44      }
45      if (targetEnum.MoveNext() | predEnum.MoveNext())
46        throw new ArgumentException("target and pred have different lengths");
47
48      return s;
49    }
50
51    public IEnumerable<double> GetLossGradient(IEnumerable<double> target, IEnumerable<double> pred) {
52      var targetEnum = target.GetEnumerator();
53      var predEnum = pred.GetEnumerator();
54
55      while (targetEnum.MoveNext() & predEnum.MoveNext()) {
56        yield return 2.0 * (targetEnum.Current - predEnum.Current); // dL(y, f(x)) / df(x)  = 2 * res
57      }
58      if (targetEnum.MoveNext() | predEnum.MoveNext())
59        throw new ArgumentException("target and pred have different lengths");
60    }
61
62    // targetArr and predArr are not changed by LineSearch
63    public double LineSearch(double[] targetArr, double[] predArr, int[] idx, int startIdx, int endIdx) {
64      if (targetArr.Length != predArr.Length)
65        throw new ArgumentException("target and pred have different lengths");
66
67      // line search for squared error loss
68      // for a given partition of rows the optimal constant that should be added to the current prediction values is the average of the residuals
69      double s = 0.0;
70      int n = 0;
71      for (int i = startIdx; i <= endIdx; i++) {
72        int row = idx[i];
73        s += (targetArr[row] - predArr[row]);
74        n++;
75      }
76      return s / n;
77    }
78
79    #region item implementation
80    private SquaredErrorLoss(SquaredErrorLoss original, Cloner cloner) : base(original, cloner) { }
81
82    public override IDeepCloneable Clone(Cloner cloner) {
83      return new SquaredErrorLoss(this, cloner);
84    }
85    #endregion
86  }
87}
Note: See TracBrowser for help on using the repository browser.