Validating cluster structures in data mining tasks