ABSTRACT
So far, we have detailed, discussed, and analyzed algorithms for several data mining tasks, with a focus on graphic data. Given a set of data and some mining goal (e.g. classification, clustering, etc.), we have a good idea of which algorithms to apply in order to analyze the data. However, this is only one half of the battle. The other half lies in understanding, (1) if the model or algorithm applied is (in)appropriate for the task at hand, and (2) if the model or algorithm produces usable results in the context of the problem. To help differentiate these outcomes, there are a number of well-established performance metrics which standardize evaluative criteria.
