|
Vol: 56(70) No: 3 / September 2011 The Decision Rules Synthesis Based on Similarity Relation Vladimir Brtka Department of Computer Science, University of Novi Sad, Technical Faculty “Mihajlo Pupin”, Djure Djakovica bb, 23000 Zrenjanin, Serbia, phone: (+381) 23 550 515, e-mail: brtkav@gmail.com Eleonora Brtka Department of Computer Science, University of Novi Sad, Technical Faculty “Mihajlo Pupin” , Djure Djakovica bb, 23000 Zrenjanin, Serbia, e-mail: brtka@sbb.rs Visnja Ognjenovic Department of Computer Science, University of Novi Sad, Technical Faculty “Mihajlo Pupin”, Djure Djakovica bb, 23000 Zrenjanin, Serbia, e-mail: visnjaognjenovic@gmail.com Ivana Berkovic Department of Computer Science, University of Novi Sad, Technical Faculty “Mihajlo Pupin”, Djure Djakovica bb, 23000 Zrenjanin, Serbia, e-mail: berki@.sbb.rs Keywords: rough sets theory, indiscernibility, similarity, decision rules Abstract This paper deals with data analyses based on the rough sets theory when table organized data is available. The application of the rough sets theory enables synthesis of the decision rules in the readable If Then form. Unlike the standard approach to decision rules synthesis when the whole data table is used, proposed approach is focused to a single element of the data table. Similarity classes are formed by similarity relation for each element. Local and global similarity relations have been described as well as the process of the aggregation of local similarities. Instead of the partition, as is the case when equivalence based indiscernibility relation is used, we have similarity classes, so decision rules are synthesized according to each similarity class separately in a standard way by rough sets toolkit Rosetta. This enables an insight to the properties of each similarity class. Possible advantage of the proposed approach is that a single element is not evaluated by its own properties only, but by the properties of entire similarity class to which the element belongs. An experiment was conducted on small data sample available via Internet ftp protocol. Generated decision rules can be analyzed by different techniques, by domain experts. Results show how the technique based the proposed approach can be implemented and used in various enterprises. References [1] Y. Zhao, Y. Yao, F. Luo,”Data Analysis Based on Discernibility and Indiscernibility”, Information Sciences, 177, Elsevier Inc., pp. 4959-4976, 2007. [2] Z. Pawlak, J. Grzymala-Busse, R. Slowinski and W. Ziarko, “Rough sets”, Association for Computing Machinery. Communications of the ACM 38, Nov 1995. [3] Z. Pawlak, A. Skowron: ”Rudiments of rough sets”, An International Journal of Information Sciences 177, 2007, pp. 3–27. [4] Z. Pawlak, “Rough set approach to knowledge-based decision support”, European Journal of Operational Research 99, 1997, pp. 48-57. [5] E.A. Abo-Tabl, “A comparison of two kinds of definitions of rough approximations based on a similarity relation”, Information Sciences 181 pp. 2587–2596, 2011. [6] V. Brtka, I. Berković, E. Stokić, B. Srdić, “A Comparison of Rule Sets Generated from Databases by Indiscernibility Relation – A Rough Sets Approach”, IEEE 3rd International Conference on Intelligent Computer Communication and Processing, Cluj-Napoca, Romania: Proceedings, 6-8 September, 2007, pp. 279–282, ISBN 1-4244-1491-1. [7] V. Brtka, E. Stokic, B. Srdic, “Automated extraction of decision rules for leptin dynamics—A rough sets approach”, Journal of Biomedical Informatics 41, 2008, pp. 667–674. [8] E. Stokic, V. Brtka, B. Srdic, “The synthesis of the rough set model for the better applicability of sagittal abdominal diameter in identifying high risk patients”, Computers in Biology and Medicine 40, 2010, pp. 786–790. [9] D. Dobrilovic, V. Brtka, I. Berkovic, B. Odadzic, “Evaluation of the Virtual Network Laboratory Exercises Using a Method Based on the Rough Set Theory”, Computer Applications in Engineering Education, Wiley Periodicals, Inc., A Wiley Company, 2009. DOI 10.1002/cae.20370, Published online in Wiley InterScience. [10] V. Brtka, I. Berkovic, E. Stokic, B. Srdic, “A Comparison of Rule Sets Generated from Databases by Indiscernibility Relation – A Rough Sets Approach”, IEEE 3rd International Conference on Intelligent Computer Communication and Processing, Cluj-Napoca, Romania: Proceedings, 2007, pp. 279–282, ISBN 1-4244-1491-1. [11] Vladimir Brtka, Ivana Berkovic, Eleonora Brtka, Vesna Jevtic, “A Comparison of Rule Sets Induced by Techniques Based on Rough Set Theory”, 6. IEEE International Symposium on Intelligent Systems and Informatics, SYSI 2008, Subotica: Library of Congress: 2008903275, IEEE Catalog Number: CFP0884C-CDR, pp 1-4, ISBN 978-1-4244-2406-1. [12] S. Greco, M. Benedetto, R. Slowinski, “New Developments in the Rough Set Approach to Multi – Attribute Decision Analysis”, Bulletin of International Rough Set Society, Volume 2, Number 2/3, September 1998; pp. 57–87. [13] A. Øhrn, “Discernibility and Rough Sets in Medicine: Tools and Applications”, Department of Computer and Information Science, Norwegian University of Science and Technology, N-7491 Trondheim, Norway, ISBN 82-7984-014-1, ISSN 0802-6394, 1999. [14] R. Slowinski, D. Vanderpooten, “A Generalized Definition of Rough Approximations Based on Similarity”, IEEE Transactions on Knowledge and Data Engineering, VOL. 12, NO. 2, MARCH/APRIL 2000. [15] S. Greco, B. Matarazzo, R. Slowinski, “Rough set theory for multicriteria decision analysis”, Eur. J. Oper. Res. 129 (1), 2001, pp. 1–47. [16] A. Liping, T. Lingyun, ”Rough approximations based on intersection of indiscernibility, similarity and outranking relations”, Knowledge-Based Systems 23, 2010, pp. 555–562. [17] L. Y. Zhai, L. P. Khoo, Z. W. Zhong, “A dominance-based rough set approach to Kansei Engineering in product development”, Expert Systems with Applications 36, 2009, pp. 393–402. [18] J.Stepaniuk, “Searching for Proper Approximation Space”, Bulletin of International Rough Sets Society, Vol. 1, Number 2, 1997, pp. 43–50. [19] K. Krawiec, R. Slowinski, D Vanderpooten, “Construction of Rough Classifiers Based on Application of a Similarity Relation”, Proceedings of the Fourth International Workshop on Rough Sets, Fuzzy Sets, and Machine Discovery, November 6-8, Tokyo, Japan, 1996, pp. 23-30. [20] D. S. Johnson, “Approximation algorithms for combinatorial problems”, Journal of Computer and System Sciences 9, 1974, pp. 256–278. |