Induction of classification rules from imperfect data

* Final gross prices may vary according to local VAT.

Get Access

Abstract

We present a method for inducing classification rules from imperfect data using an extended version of the rough set model. The salient feature of our method is that it makes use of the statistical information inherent in the information system. Our framework describes the overall induction task in terms of two key subtasks: approximate classification and rule generation.