Rough set-based approach for automated discovery of censored production rules

Abstract

This paper proposes an algorithm based on rough set (RS) framework for automated discovery of censored production rules with provided operator. These censored production rules are created by augmenting ordinary production rules with rare or particular conditions employing unless operator or provided operator, respectively. Systems which use censored production rules are free to ignore the censor conditions when the resources are limited. Censor conditions can be examined later leading to more specific and accurate rules if the time permits or resources are at hand. Integration of the RS theory with censored production rules allows for handling of vague and imprecise information in a limited resource environment. The proposed algorithm generates a separate list of rules for the lower and upper approximation of the concept. The rule belonging to lower approximation is a certain rule, and the rule belonging to upper approximation is a possible rule, which can be further investigated using the censor conditions. We extend the basic notion of production rules using the censor condition with provided operator to find out the certain rules in the given data-set. We have investigated our algorithm on the benchmark data-sets from the UCI machine learning repository, which have shown promising results.