public Experiment(string name, EEGRecord er, AMLearning ml, bool enabled)
{
Name = name;
_er = er;
_ml = ml;
_enabled = enabled;
_ml.Progress += new ChangedValuesEventHandler(_ml_Progress);
}
public void SaveModel(AMLearning model)
{
if (!File.Exists(DbSettings.fullpath + ".eq"))
{
db.CreateDatabase(DbSettings.fullpath);
}
db.OpenDatabase(DbSettings.fullpath);
db.RefreshMode = ObjectRefreshMode.AlwaysReturnUpdatedValues;
db.Store(model);
db.Close();
}
private void listBoxModels_SelectedIndexChanged(object sender, EventArgs e)
{
if (listBoxModels.SelectedIndex != -1)
{
model = models[listBoxModels.SelectedIndex];
listBoxClasses.Items.Clear();
foreach (var item in model.ActionList)
{
listBoxClasses.Items.Add(item.Key);
}
}
}
private void buttonCalculate_Click(object sender, EventArgs e)
{
//if (AsyncWorkerCalculate.IsBusy)
//{
// buttonCalculate.Enabled = false;
// AsyncWorkerCalculate.CancelAsync();
//}
//else //start new process
{
listBoxLogger.Items.Insert(0, "Creating machine learning model to be used for classification...");
if (currentRecord.FeatureVectorsOutputInput.Count == 0)
{
MessageBox.Show("First you need to record/load some data for specific action!"); return;
}
buttonCalculate.Enabled = false;
//buttonCalculate.Text = "Cancel";
startCalculateModel = DateTime.Now;
switch (cbMethods.SelectedIndex)
{
case 0: model = new LdaSVM("ldasvm"); break;
case 1: model = new LdaMLP("ldamlp"); break;
case 2: model = new LdaRBF("ldarbf"); break;
case 3: model = new OctaveMulticlassLogisticRegression("omlr"); break;
}
model.ActionList = currentRecord.actions;
model.Progress += new ChangedValuesEventHandler(model_Progress);
AsyncWorkerCalculate.RunWorkerAsync();
}
}
