This FS scheme is formed by two steps: a) an incremental feature ranking method, and b) an incremental learning algorithm that can consider a subset of the features during prediction (Naive Bayes).
val gain = InfoGainTransformer() .setNFeatures(2) .setSelectNF(1) gain fit dataSet val result = gain transform dataSet
I. Katakis, G. Tsoumakas, I. Vlahavas, Advances in Informatics: 10th Panhellenic Conference on Informatics, PCI 2005, Springer Berlin Heidelberg, 2005, Ch. On the Utility of Incremental Feature Selection for the Classification of Textual Data Streams, pp. 338–348.