Version md: decision tree with extra class

[educsantos]

The algorithm is always expecting one more class. For instance, the iris dataset contains three different species, but the algorithm reports, in the table of parameters and in the tree nodes, that there are four classes. The same happened with other datasets.