Rule Generation and Compactation in the WWTP

Abstract

Abstract. In this paper we discuss our approach to learning c1assification rules from data. We sketch out two modules ofour architecture, namely LINNEO+ and GAR.LINNEO+, which is a knowledge acquisition tool for ill-structured domains automatically generating cJasses from examples that incrementally works with an unsupervised strategy. LINNEO+'s output, a representation of the conceptual structure ofthe domain in terms of cJasses, is the input to GAR that is used to generate a set of classification rules for the original training set. GAR can generate both conjuctive and disjunctive rules. Herein we present an application of these techniques to data obtained from a real wastewater treatment plant in order to help the construction of a rule base. This rule wilJ be used for a knowledge-based system that aims to supervise the whole process,