I have almost finished converting the Backpropagation to a multi-class algorithm. As of now, it is a binary learner. There had been notes left in learning.py about the issue (approximately lines 545 and 650).

With the BackPropagationLearner now multi-class, Neural Networks and the Perceptron algorithms now can take care of multi-class datasets.

PS: I will probably wait till some crucial PRs are merged before pulling this one, since this change requires editing across the learning module.