public Benchmark(string name, IEnumerable<Learner> learners, IEnumerable<int> dataSets, int trainingSetSize, int validationSetSize, int numberOfRuns, bool useTestData = false)
{
Name = name;
this.dataSets = dataSets.Select(num => new DataSet(num, trainingSetSize, validationSetSize)).ToArray();
this.numberOfRuns = numberOfRuns;
this.learners = new Dictionary<Learner, LearnerParameters>();
this.useTestData = useTestData;
foreach (Learner learner in learners)
{
LearnerParameters parameters = new LearnerParameters(learner.Name());
this.learners.Add(learner, parameters);
}
}
public abstract void Initialise(LearnerParameters parameters, int iteration);
private string IterationName(LearnerParameters parameters, int iteration)
{
if (parameters.StepSize == 0)
{
return "Static Setting";
}
return String.Format("{0}: {1}", parameters.RunningParameterName, (parameters.Minimum + (iteration * parameters.StepSize)).ToString(parameters.RunningParameterDecimal ? "00.000000" : "000"));
}
public override void Initialise(LearnerParameters parameters,
int iteration) {
maximum_iterations = parameters.MaxIterations;
maximum_states = parameters.MaxStates;
//const int NUM_SYMBOLS = 42;
//HMMGraph graph = new HMMGraph(NUM_SYMBOLS);
//for (int i = 0; i < parameters.Minimum; i++) {
// graph.AddNode(new Node());
//}
//foreach (Node n in graph.Nodes) {
// foreach (Node m in graph.Nodes) {
// n.SetTransition(m, 0.5);
// }
// for (int i = 0; i < NUM_SYMBOLS; i++) {
// n.SetEmission(i, 0.5);
// }
//}
//graph.Normalize();
//this.hmm = SparseHiddenMarkovModel.FromGraph(graph);
}