Free cookie consent management tool by TermsFeed Policy Generator

source: trunk/sources/HeuristicLab.Algorithms.DataAnalysis/3.4/GBM/GradientBoostingRegressionAlgorithm.cs @ 14064

Last change on this file since 14064 was 13978, checked in by gkronber, 8 years ago

#1795: removed MCTS symb reg as algorithm in GBM to cut the dependency between GBM and MCTS symb reg (MCTS symb reg not ready for release yet)

File size: 22.8 KB
RevLine 
[13646]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 System.Threading;
27using HeuristicLab.Analysis;
28using HeuristicLab.Common;
29using HeuristicLab.Core;
30using HeuristicLab.Data;
31using HeuristicLab.Encodings.SymbolicExpressionTreeEncoding;
32using HeuristicLab.Optimization;
33using HeuristicLab.Parameters;
34using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
35using HeuristicLab.Problems.DataAnalysis;
36using HeuristicLab.Problems.DataAnalysis.Symbolic;
37using HeuristicLab.Problems.DataAnalysis.Symbolic.Regression;
38using HeuristicLab.Random;
[13653]39using HeuristicLab.Selection;
[13646]40
41namespace HeuristicLab.Algorithms.DataAnalysis.MctsSymbolicRegression {
42  [Item("Gradient Boosting Machine Regression (GBM)",
43    "Gradient boosting for any regression base learner (e.g. MCTS symbolic regression)")]
44  [StorableClass]
45  [Creatable(CreatableAttribute.Categories.DataAnalysisRegression, Priority = 350)]
46  public class GradientBoostingRegressionAlgorithm : BasicAlgorithm {
47    public override Type ProblemType {
48      get { return typeof(IRegressionProblem); }
49    }
50
51    public new IRegressionProblem Problem {
52      get { return (IRegressionProblem)base.Problem; }
53      set { base.Problem = value; }
54    }
55
56    #region ParameterNames
57
58    private const string IterationsParameterName = "Iterations";
59    private const string NuParameterName = "Nu";
60    private const string MParameterName = "M";
61    private const string RParameterName = "R";
62    private const string RegressionAlgorithmParameterName = "RegressionAlgorithm";
63    private const string SeedParameterName = "Seed";
64    private const string SetSeedRandomlyParameterName = "SetSeedRandomly";
65    private const string CreateSolutionParameterName = "CreateSolution";
[13889]66    private const string StoreRunsParameterName = "StoreRuns";
[13646]67    private const string RegressionAlgorithmSolutionResultParameterName = "RegressionAlgorithmResult";
68
69    #endregion
70
71    #region ParameterProperties
72
73    public IFixedValueParameter<IntValue> IterationsParameter {
74      get { return (IFixedValueParameter<IntValue>)Parameters[IterationsParameterName]; }
75    }
76
77    public IFixedValueParameter<DoubleValue> NuParameter {
78      get { return (IFixedValueParameter<DoubleValue>)Parameters[NuParameterName]; }
79    }
80
81    public IFixedValueParameter<DoubleValue> RParameter {
82      get { return (IFixedValueParameter<DoubleValue>)Parameters[RParameterName]; }
83    }
84
85    public IFixedValueParameter<DoubleValue> MParameter {
86      get { return (IFixedValueParameter<DoubleValue>)Parameters[MParameterName]; }
87    }
88
89    // regression algorithms are currently: DataAnalysisAlgorithms, BasicAlgorithms and engine algorithms with no common interface
90    public IConstrainedValueParameter<IAlgorithm> RegressionAlgorithmParameter {
91      get { return (IConstrainedValueParameter<IAlgorithm>)Parameters[RegressionAlgorithmParameterName]; }
92    }
93
94    public IFixedValueParameter<StringValue> RegressionAlgorithmSolutionResultParameter {
95      get { return (IFixedValueParameter<StringValue>)Parameters[RegressionAlgorithmSolutionResultParameterName]; }
96    }
97
98    public IFixedValueParameter<IntValue> SeedParameter {
99      get { return (IFixedValueParameter<IntValue>)Parameters[SeedParameterName]; }
100    }
101
102    public FixedValueParameter<BoolValue> SetSeedRandomlyParameter {
103      get { return (FixedValueParameter<BoolValue>)Parameters[SetSeedRandomlyParameterName]; }
104    }
105
106    public IFixedValueParameter<BoolValue> CreateSolutionParameter {
107      get { return (IFixedValueParameter<BoolValue>)Parameters[CreateSolutionParameterName]; }
108    }
[13889]109    public IFixedValueParameter<BoolValue> StoreRunsParameter {
110      get { return (IFixedValueParameter<BoolValue>)Parameters[StoreRunsParameterName]; }
111    }
[13646]112
113    #endregion
114
115    #region Properties
116
117    public int Iterations {
118      get { return IterationsParameter.Value.Value; }
119      set { IterationsParameter.Value.Value = value; }
120    }
121
122    public int Seed {
123      get { return SeedParameter.Value.Value; }
124      set { SeedParameter.Value.Value = value; }
125    }
126
127    public bool SetSeedRandomly {
128      get { return SetSeedRandomlyParameter.Value.Value; }
129      set { SetSeedRandomlyParameter.Value.Value = value; }
130    }
131
132    public double Nu {
133      get { return NuParameter.Value.Value; }
134      set { NuParameter.Value.Value = value; }
135    }
136
137    public double R {
138      get { return RParameter.Value.Value; }
139      set { RParameter.Value.Value = value; }
140    }
141
142    public double M {
143      get { return MParameter.Value.Value; }
144      set { MParameter.Value.Value = value; }
145    }
146
147    public bool CreateSolution {
148      get { return CreateSolutionParameter.Value.Value; }
149      set { CreateSolutionParameter.Value.Value = value; }
150    }
151
[13889]152    public bool StoreRuns {
153      get { return StoreRunsParameter.Value.Value; }
154      set { StoreRunsParameter.Value.Value = value; }
155    }
156
[13646]157    public IAlgorithm RegressionAlgorithm {
158      get { return RegressionAlgorithmParameter.Value; }
159    }
160
161    public string RegressionAlgorithmResult {
162      get { return RegressionAlgorithmSolutionResultParameter.Value.Value; }
163      set { RegressionAlgorithmSolutionResultParameter.Value.Value = value; }
164    }
165
166    #endregion
167
168    [StorableConstructor]
169    protected GradientBoostingRegressionAlgorithm(bool deserializing)
170      : base(deserializing) {
171    }
172
173    protected GradientBoostingRegressionAlgorithm(GradientBoostingRegressionAlgorithm original, Cloner cloner)
174      : base(original, cloner) {
175    }
176
177    public override IDeepCloneable Clone(Cloner cloner) {
178      return new GradientBoostingRegressionAlgorithm(this, cloner);
179    }
180
181    public GradientBoostingRegressionAlgorithm() {
182      Problem = new RegressionProblem(); // default problem
[13978]183      var osgp = CreateOSGP();
[13646]184      var regressionAlgs = new ItemSet<IAlgorithm>(new IAlgorithm[] {
[13653]185        new RandomForestRegression(),
[13978]186        osgp,
[13646]187      });
188      foreach (var alg in regressionAlgs) alg.Prepare();
189
190
191      Parameters.Add(new FixedValueParameter<IntValue>(IterationsParameterName,
192        "Number of iterations", new IntValue(100)));
193      Parameters.Add(new FixedValueParameter<IntValue>(SeedParameterName,
194        "The random seed used to initialize the new pseudo random number generator.", new IntValue(0)));
195      Parameters.Add(new FixedValueParameter<BoolValue>(SetSeedRandomlyParameterName,
196        "True if the random seed should be set to a random value, otherwise false.", new BoolValue(true)));
197      Parameters.Add(new FixedValueParameter<DoubleValue>(NuParameterName,
198        "The learning rate nu when updating predictions in GBM (0 < nu <= 1)", new DoubleValue(0.5)));
199      Parameters.Add(new FixedValueParameter<DoubleValue>(RParameterName,
200        "The fraction of rows that are sampled randomly for the base learner in each iteration (0 < r <= 1)",
201        new DoubleValue(1)));
202      Parameters.Add(new FixedValueParameter<DoubleValue>(MParameterName,
203        "The fraction of variables that are sampled randomly for the base learner in each iteration (0 < m <= 1)",
204        new DoubleValue(0.5)));
205      Parameters.Add(new ConstrainedValueParameter<IAlgorithm>(RegressionAlgorithmParameterName,
[13978]206        "The regression algorithm to use as a base learner", regressionAlgs, osgp));
[13646]207      Parameters.Add(new FixedValueParameter<StringValue>(RegressionAlgorithmSolutionResultParameterName,
208        "The name of the solution produced by the regression algorithm", new StringValue("Solution")));
209      Parameters[RegressionAlgorithmSolutionResultParameterName].Hidden = true;
210      Parameters.Add(new FixedValueParameter<BoolValue>(CreateSolutionParameterName,
211        "Flag that indicates if a solution should be produced at the end of the run", new BoolValue(true)));
212      Parameters[CreateSolutionParameterName].Hidden = true;
[13889]213      Parameters.Add(new FixedValueParameter<BoolValue>(StoreRunsParameterName,
214        "Flag that indicates if the results of the individual runs should be stored for detailed analysis", new BoolValue(false)));
215      Parameters[StoreRunsParameterName].Hidden = true;
[13646]216    }
217
218    protected override void Run(CancellationToken cancellationToken) {
219      // Set up the algorithm
220      if (SetSeedRandomly) Seed = new System.Random().Next();
221      var rand = new MersenneTwister((uint)Seed);
222
223      // Set up the results display
224      var iterations = new IntValue(0);
225      Results.Add(new Result("Iterations", iterations));
226
227      var table = new DataTable("Qualities");
[13889]228      table.Rows.Add(new DataRow("R² (train)"));
229      table.Rows.Add(new DataRow("R² (test)"));
[13646]230      Results.Add(new Result("Qualities", table));
231      var curLoss = new DoubleValue();
232      var curTestLoss = new DoubleValue();
[13889]233      Results.Add(new Result("R² (train)", curLoss));
234      Results.Add(new Result("R² (test)", curTestLoss));
[13646]235      var runCollection = new RunCollection();
[13889]236      if (StoreRuns)
237        Results.Add(new Result("Runs", runCollection));
[13646]238
239      // init
240      var problemData = Problem.ProblemData;
[13889]241      var targetVarName = problemData.TargetVariable;
[13707]242      var activeVariables = problemData.AllowedInputVariables.Concat(new string[] { problemData.TargetVariable });
[13646]243      var modifiableDataset = new ModifiableDataset(
[13707]244        activeVariables,
245        activeVariables.Select(v => problemData.Dataset.GetDoubleValues(v).ToList()));
[13646]246
247      var trainingRows = problemData.TrainingIndices;
248      var testRows = problemData.TestIndices;
249      var yPred = new double[trainingRows.Count()];
250      var yPredTest = new double[testRows.Count()];
251      var y = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, problemData.TrainingIndices).ToArray();
252      var curY = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, problemData.TrainingIndices).ToArray();
253
254      var yTest = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, problemData.TestIndices).ToArray();
255      var curYTest = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, problemData.TestIndices).ToArray();
256      var nu = Nu;
257      var mVars = (int)Math.Ceiling(M * problemData.AllowedInputVariables.Count());
258      var rRows = (int)Math.Ceiling(R * problemData.TrainingIndices.Count());
259      var alg = RegressionAlgorithm;
260      List<IRegressionModel> models = new List<IRegressionModel>();
261      try {
262
263        // Loop until iteration limit reached or canceled.
264        for (int i = 0; i < Iterations; i++) {
265          cancellationToken.ThrowIfCancellationRequested();
266
267          modifiableDataset.RemoveVariable(targetVarName);
268          modifiableDataset.AddVariable(targetVarName, curY.Concat(curYTest));
269
270          SampleTrainingData(rand, modifiableDataset, rRows, problemData.Dataset, curY, problemData.TargetVariable, problemData.TrainingIndices); // all training indices from the original problem data are allowed
271          var modifiableProblemData = new RegressionProblemData(modifiableDataset,
272            problemData.AllowedInputVariables.SampleRandomWithoutRepetition(rand, mVars),
273            problemData.TargetVariable);
274          modifiableProblemData.TrainingPartition.Start = 0;
275          modifiableProblemData.TrainingPartition.End = rRows;
276          modifiableProblemData.TestPartition.Start = problemData.TestPartition.Start;
277          modifiableProblemData.TestPartition.End = problemData.TestPartition.End;
278
279          if (!TrySetProblemData(alg, modifiableProblemData))
280            throw new NotSupportedException("The algorithm cannot be used with GBM.");
281
282          IRegressionModel model;
283          IRun run;
[13898]284
[13646]285          // try to find a model. The algorithm might fail to produce a model. In this case we just retry until the iterations are exhausted
[13898]286          if (TryExecute(alg, rand.Next(), RegressionAlgorithmResult, out model, out run)) {
[13646]287            int row = 0;
288            // update predictions for training and test
289            // update new targets (in the case of squared error loss we simply use negative residuals)
290            foreach (var pred in model.GetEstimatedValues(problemData.Dataset, trainingRows)) {
291              yPred[row] = yPred[row] + nu * pred;
292              curY[row] = y[row] - yPred[row];
293              row++;
294            }
295            row = 0;
296            foreach (var pred in model.GetEstimatedValues(problemData.Dataset, testRows)) {
297              yPredTest[row] = yPredTest[row] + nu * pred;
298              curYTest[row] = yTest[row] - yPredTest[row];
299              row++;
300            }
301            // determine quality
302            OnlineCalculatorError error;
303            var trainR = OnlinePearsonsRCalculator.Calculate(yPred, y, out error);
304            var testR = OnlinePearsonsRCalculator.Calculate(yPredTest, yTest, out error);
305
306            // iteration results
307            curLoss.Value = error == OnlineCalculatorError.None ? trainR * trainR : 0.0;
308            curTestLoss.Value = error == OnlineCalculatorError.None ? testR * testR : 0.0;
309
310            models.Add(model);
311
312
313          }
314
[13889]315          if (StoreRuns)
316            runCollection.Add(run);
317          table.Rows["R² (train)"].Values.Add(curLoss.Value);
318          table.Rows["R² (test)"].Values.Add(curTestLoss.Value);
[13646]319          iterations.Value = i + 1;
320        }
321
322        // produce solution
323        if (CreateSolution) {
324          // when all our models are symbolic models we can easily combine them to a single model
325          if (models.All(m => m is ISymbolicRegressionModel)) {
326            Results.Add(new Result("Solution", CreateSymbolicSolution(models, Nu, (IRegressionProblemData)problemData.Clone())));
327          }
328          // just produce an ensemble solution for now (TODO: correct scaling or linear regression for ensemble model weights)
[13699]329
[13917]330          var ensembleSolution = CreateEnsembleSolution(models, (IRegressionProblemData)problemData.Clone());
[13699]331          Results.Add(new Result("EnsembleSolution", ensembleSolution));
[13646]332        }
[13699]333      }
334      finally {
[13646]335        // reset everything
336        alg.Prepare(true);
337      }
338    }
339
[13917]340    private static IRegressionEnsembleSolution CreateEnsembleSolution(List<IRegressionModel> models,
341      IRegressionProblemData problemData) {
342      var rows = problemData.TrainingPartition.Size;
343      var features = models.Count;
344      double[,] inputMatrix = new double[rows, features + 1];
345      //add model estimates
346      for (int m = 0; m < models.Count; m++) {
347        var model = models[m];
348        var estimates = model.GetEstimatedValues(problemData.Dataset, problemData.TrainingIndices);
349        int estimatesCounter = 0;
350        foreach (var estimate in estimates) {
351          inputMatrix[estimatesCounter, m] = estimate;
352          estimatesCounter++;
353        }
354      }
[13653]355
[13917]356      //add target
357      var targets = problemData.Dataset.GetDoubleValues(problemData.TargetVariable, problemData.TrainingIndices);
358      int targetCounter = 0;
359      foreach (var target in targets) {
360        inputMatrix[targetCounter, models.Count] = target;
361        targetCounter++;
362      }
363
364      alglib.linearmodel lm = new alglib.linearmodel();
365      alglib.lrreport ar = new alglib.lrreport();
366      double[] coefficients;
367      int retVal = 1;
368      alglib.lrbuildz(inputMatrix, rows, features, out retVal, out lm, out ar);
369      if (retVal != 1) throw new ArgumentException("Error in calculation of linear regression solution");
370
371      alglib.lrunpack(lm, out coefficients, out features);
372
373      var ensembleModel = new RegressionEnsembleModel(models, coefficients.Take(models.Count)) { AverageModelEstimates = false };
[13941]374      var ensembleSolution = (IRegressionEnsembleSolution)ensembleModel.CreateRegressionSolution(problemData);
[13917]375      return ensembleSolution;
376    }
377
378
[13653]379    private IAlgorithm CreateOSGP() {
380      // configure strict osgp
381      var alg = new OffspringSelectionGeneticAlgorithm.OffspringSelectionGeneticAlgorithm();
382      var prob = new SymbolicRegressionSingleObjectiveProblem();
383      prob.MaximumSymbolicExpressionTreeDepth.Value = 7;
384      prob.MaximumSymbolicExpressionTreeLength.Value = 15;
385      alg.Problem = prob;
386      alg.SuccessRatio.Value = 1.0;
387      alg.ComparisonFactorLowerBound.Value = 1.0;
388      alg.ComparisonFactorUpperBound.Value = 1.0;
389      alg.MutationProbability.Value = 0.15;
390      alg.PopulationSize.Value = 200;
391      alg.MaximumSelectionPressure.Value = 100;
392      alg.MaximumEvaluatedSolutions.Value = 20000;
393      alg.SelectorParameter.Value = alg.SelectorParameter.ValidValues.OfType<GenderSpecificSelector>().First();
394      alg.MutatorParameter.Value = alg.MutatorParameter.ValidValues.OfType<MultiSymbolicExpressionTreeManipulator>().First();
395      alg.StoreAlgorithmInEachRun = false;
396      return alg;
397    }
398
[13646]399    private void SampleTrainingData(MersenneTwister rand, ModifiableDataset ds, int rRows,
400      IDataset sourceDs, double[] curTarget, string targetVarName, IEnumerable<int> trainingIndices) {
401      var selectedRows = trainingIndices.SampleRandomWithoutRepetition(rand, rRows).ToArray();
402      int t = 0;
403      object[] srcRow = new object[ds.Columns];
404      var varNames = ds.DoubleVariables.ToArray();
405      foreach (var r in selectedRows) {
406        // take all values from the original dataset
407        for (int c = 0; c < srcRow.Length; c++) {
408          var col = sourceDs.GetReadOnlyDoubleValues(varNames[c]);
409          srcRow[c] = col[r];
410        }
411        ds.ReplaceRow(t, srcRow);
412        // but use the updated target values
413        ds.SetVariableValue(curTarget[r], targetVarName, t);
414        t++;
415      }
416    }
417
418    private static ISymbolicRegressionSolution CreateSymbolicSolution(List<IRegressionModel> models, double nu, IRegressionProblemData problemData) {
419      var symbModels = models.OfType<ISymbolicRegressionModel>();
420      var lowerLimit = symbModels.Min(m => m.LowerEstimationLimit);
421      var upperLimit = symbModels.Max(m => m.UpperEstimationLimit);
422      var interpreter = new SymbolicDataAnalysisExpressionTreeLinearInterpreter();
423      var progRootNode = new ProgramRootSymbol().CreateTreeNode();
424      var startNode = new StartSymbol().CreateTreeNode();
425
426      var addNode = new Addition().CreateTreeNode();
427      var mulNode = new Multiplication().CreateTreeNode();
428      var scaleNode = (ConstantTreeNode)new Constant().CreateTreeNode(); // all models are scaled using the same nu
429      scaleNode.Value = nu;
430
431      foreach (var m in symbModels) {
432        var relevantPart = m.SymbolicExpressionTree.Root.GetSubtree(0).GetSubtree(0); // skip root and start
433        addNode.AddSubtree((ISymbolicExpressionTreeNode)relevantPart.Clone());
434      }
435
436      mulNode.AddSubtree(addNode);
437      mulNode.AddSubtree(scaleNode);
438      startNode.AddSubtree(mulNode);
439      progRootNode.AddSubtree(startNode);
440      var t = new SymbolicExpressionTree(progRootNode);
[13941]441      var combinedModel = new SymbolicRegressionModel(problemData.TargetVariable, t, interpreter, lowerLimit, upperLimit);
[13646]442      var sol = new SymbolicRegressionSolution(combinedModel, problemData);
443      return sol;
444    }
445
446    private static bool TrySetProblemData(IAlgorithm alg, IRegressionProblemData problemData) {
447      var prob = alg.Problem as IRegressionProblem;
448      // there is already a problem and it is compatible -> just set problem data
449      if (prob != null) {
450        prob.ProblemDataParameter.Value = problemData;
451        return true;
[13653]452      } else return false;
[13646]453    }
454
[13898]455    private static bool TryExecute(IAlgorithm alg, int seed, string regressionAlgorithmResultName, out IRegressionModel model, out IRun run) {
[13646]456      model = null;
[13898]457      SetSeed(alg, seed);
[13646]458      using (var wh = new AutoResetEvent(false)) {
[13889]459        Exception ex = null;
460        EventHandler<EventArgs<Exception>> handler = (sender, args) => {
461          ex = args.Value;
462          wh.Set();
463        };
[13646]464        EventHandler handler2 = (sender, args) => wh.Set();
465        alg.ExceptionOccurred += handler;
466        alg.Stopped += handler2;
467        try {
468          alg.Prepare();
469          alg.Start();
470          wh.WaitOne();
471
[13889]472          if (ex != null) throw new AggregateException(ex);
[13646]473          run = alg.Runs.Last();
[13889]474          alg.Runs.Clear();
[13646]475          var sols = alg.Results.Select(r => r.Value).OfType<IRegressionSolution>();
476          if (!sols.Any()) return false;
477          var sol = sols.First();
478          if (sols.Skip(1).Any()) {
479            // more than one solution => use regressionAlgorithmResult
480            if (alg.Results.ContainsKey(regressionAlgorithmResultName)) {
481              sol = (IRegressionSolution)alg.Results[regressionAlgorithmResultName].Value;
482            }
483          }
484          var symbRegSol = sol as SymbolicRegressionSolution;
485          // only accept symb reg solutions that do not hit the estimation limits
486          // NaN evaluations would not be critical but are problematic if we want to combine all symbolic models into a single symbolic model
487          if (symbRegSol == null ||
488            (symbRegSol.TrainingLowerEstimationLimitHits == 0 && symbRegSol.TrainingUpperEstimationLimitHits == 0 &&
489             symbRegSol.TestLowerEstimationLimitHits == 0 && symbRegSol.TestUpperEstimationLimitHits == 0) &&
490            symbRegSol.TrainingNaNEvaluations == 0 && symbRegSol.TestNaNEvaluations == 0) {
491            model = sol.Model;
492          }
[13699]493        }
494        finally {
[13646]495          alg.ExceptionOccurred -= handler;
496          alg.Stopped -= handler2;
497        }
498      }
499      return model != null;
500    }
[13898]501
502    private static void SetSeed(IAlgorithm alg, int seed) {
503      // no common interface for algs that use a PRNG -> use naming convention to set seed
504      var paramItem = alg as IParameterizedItem;
505
506      if (paramItem.Parameters.ContainsKey("SetSeedRandomly")) {
507        ((BoolValue)paramItem.Parameters["SetSeedRandomly"].ActualValue).Value = false;
508        ((IntValue)paramItem.Parameters["Seed"].ActualValue).Value = seed;
509      } else {
510        throw new ArgumentException("Base learner does not have a seed parameter (algorithm {0})", alg.Name);
511      }
512
513    }
[13646]514  }
515}
Note: See TracBrowser for help on using the repository browser.