Title: An evolutionary approach to discover intra- and inter-class exceptions in databases
Authors: Jyoti Vashishtha; Dharminder Kumar; Saroj Ratnoo
Addresses: Department of Computer Science & Engineering, Guru Jambheshwar University of Science & Technology, Hisar, Haryana, India ' Department of Computer Science & Engineering, Guru Jambheshwar University of Science & Technology, Hisar, Haryana, India ' Department of Computer Science & Engineering, Guru Jambheshwar University of Science & Technology, Hisar, Haryana, India
Abstract: Data mining algorithms produce information of a statistical nature that contains accurate and reliable knowledge. However, in many cases these algorithms do not discover hidden facts which may be interesting to users. Therefore, the recent aim of knowledge discovery in databases (KDD) is to expose patterns that are exceptions to the existing knowledge. Exceptions are considered interesting as they add extraordinary facts to the knowledge base and create an incentive in users to re-examine their decisions. Discovering exceptions along with the decision rules increases the quality of decision making in those rare circumstances where the rules are not applicable. This paper is an attempt to devise a framework to discover exceptions using an evolutionary approach. In this work, we have categorised exceptions as inter and intra class. Experimental results are presented to demonstrate the performance of the proposed algorithm.
Keywords: data mining; rule discovery; interesting rules; exception discovery; intra class exceptions; inter class exceptions; evolutionary approach; genetic algorithms; knowledge discovery.
DOI: 10.1504/IJISTA.2013.056535
International Journal of Intelligent Systems Technologies and Applications, 2013 Vol.12 No.3/4, pp.283 - 300
Received: 30 Nov 2012
Accepted: 13 May 2013
Published online: 25 Sep 2013 *