Discovery of Exceptions: A Step towards Perfection.
ABSTRACT It is interesting to discover exceptions, as they dispute the existing knowledge and have elements of unexpectedness and surprise. As exceptions focus on a very small portion of data, discovering exceptions still remains a great challenge. A censored production rule (CPR) is a special kind of knowledge structure that augments exceptions to their corresponding commonsense rules of high generality and support. This paper proposes discovery of decision rules in the form of censored production rules by employing a genetic algorithm approach. Results confirm that the proposed discovery of decision rules in the form of CPRs is comprehensible and interesting. Using CPRs as underlying knowledge structure for rule mining provides an excellent mechanism for exception handling and approximate reasoning. Moreover, discovering exceptions through CPRs enhances the predictive accuracy of the classifier.
- [Show abstract] [Hide abstract]
ABSTRACT: Variable precision logic offers mechanisms for handling trade-offs between the precision of inferences and the computational efficiency of deriving them. In this paper we present a Hierarchical Censored Production Rules (HCPRs) system as an underlying representational and computational scheme for variable precision logic. The proposed scheme is based on an extension of Production Rules system.It is shown how an ordinary Production Rule on suitable modification and augmentation with relevant information becomes a Hierarchical Censored Production Rule (HCPR), which in turn enables to resolve many of the problems associated with usual Production Rules system.A concept of Hierarchical Censored Production Rule tree (HCPR-tree) is developed and it is shown that it can be employed to exhibit both variable specificity and certainty of belief in a conclusion using a General Control Scheme (GCS).Data & Knowledge Engineering 04/1992; 8(1):19–34. · 1.49 Impact Factor
- Communications of the ACM 01/1996; 39:27-34. · 2.86 Impact Factor
Conference Paper: "Rule + Exception" Strategies for Knowledge Management and Discovery.[Show abstract] [Hide abstract]
ABSTRACT: A common practice of human learning and knowledge management is to use general rules, exception rules, and exceptions to rules. One of the crucial issues is to find a right mixture of them. For discovering this type of knowledge, we consider “rule + exception”, or rule-plus-exception, strategies. Results from psychology, expert systems, genetic algorithms, and machine learning and data mining are summarized and compared, and their implications to knowledge management and discovery are examined. The study motivates and establishes a basis for the design and implementation of new algorithms for the discovery of “rule + exception” type knowledge.Rough Sets, Fuzzy Sets, Data Mining, and Granular Computing, 10th International Conference, RSFDGrC 2005, Regina, Canada, August 31 - September 3, 2005, Proceedings, Part II; 01/2005