public void Expand()
{
LearningOpenNode node = _tree.OpenNode() as LearningOpenNode;
if (node == null)
{
throw new Exception("No open node left");
}
ItemSet set = node.LearningSet();
double entropy = set.CalEntropy(_goalAttribute);
if (entropy <= _entropyThreshold || _testAttributeSet.Size() == 0)
{
MakeLeafNode(node);
}
else
{
TestScore testScore = set.BestSplitTest(_testAttributeSet, _goalAttribute);
if (testScore.Score * set.Size() <= _scoreThreshold)
{ //forward pruning: test does not provide enough information
MakeLeafNode(node);
}
else
{
MakeTestNode(node, testScore.Test, testScore.Score * set.Size());
}
}
}
/// <summary>
/// Finds the test on each attribute performing the best split for finding the value of a 'goal'
/// attribute.
/// </summary>
/// <param name="candidateAttributes"></param>
/// <param name="goalAttribute"></param>
/// <returns></returns>
public IEnumerable <TestScore> BestSplitTests(AttributeSet candidateAttributes, SymbolicAttribute goalAttribute)
{
if (candidateAttributes == null || goalAttribute == null || candidateAttributes.Size() == 0)
{
throw new ArgumentNullException();
}
List <TestScore> bestScores = new List <TestScore>();
List <Attribute> attributes = candidateAttributes.GetAttributes().ToList();
foreach (Attribute attr in attributes)
{
bestScores.Add(BestSplitTest(attr, goalAttribute));
}
return(bestScores);
}