Partial classification using association rules

Abstract

Many real-life problems require a partial classification of the data. We use the term "partial classification" to describe the discovery of models that show characteristics of the data classes, but may not cover all classes and all examples of any given class. Complete classification may be infeasible or undesirable when there are a very large number of class attributes, most attributes values are missing, or the class distribution is highly skewed and the user is interested in understanding the low-frequency class. We show how association rules can be used for partial classification in such domains, and present two case studies: reducing telecommunications order failures and detecting redundant medical tests. Introduction Classification, or supervised learning, has been widely studied in the machine learning community, e.g. (Michie, Spiegelhalter, & Taylor 1994). The input data for classification, also called the training set, consists of multiple examples (records), each having mult...