Discriminators for use in flow-based classification

摘要: Any assessment of classification techniques requires data. This document describes sets data intended to aid in the work. A number are described; each set consists a objects, and object is described by group features (also referred as discriminators). Leveraged quantity hand-classified data, within represents single flow TCP packets between client server. The for consist (application-centric) derived elsewhere input probabilistic techniques. In addition describing features, we also provide information allowing interested parties retrieve these use their own contain no site-identifying information; only statistics class that defines causal application.