Hi all, Is there any existing LabeledVectorWritable implementation to combine a possibly empty list of label indexes and a vector of features to process and store labeled datasets for supervised learning algorithms?
Maybe it would be worth having a two versions of such as Writable, one for singly classified documents (one and only one label index perf instance) and another for multi categorized documents (0 or more category indexes per instance). -- Olivier http://twitter.com/ogrisel - http://code.oliviergrisel.name
