Node that oversample patterns in an input table.

The algorithm works roughly as follows: It oversamples patterns of a given class by drawing random representatives from the input table, calculating their kNNs and then choosing any nearest neighbor and putting the new pattern along the line between the representative and its nearest neighbor.

The implementation is based on the SMOTE algorithm:

		Chawla, N.V., Bowyer, K.W., Hall, L.O., Kegelmeyer, W.P. (2002) 
		"SMOTE: Synthetic Minority Over-sampling Technique",
		Journal of Artificial Intelligence Research, Volume 16, pages 321-357.