Attribute Coding for the Rough Set Theory Based Rule Simplications by Using the Particle Swarm Optimization Algorithm
The attribute coding approach has been used in the Rough Set Theory (RST) based classification problems. The attribute coding defined ranges of the attribute values as multi-thresholds. If attribute values can be defined as appropriate values, the appropriate number of rules will be generated. The attribute coding for the RST based rule derivations significantly reduces unnecessary rules and simplifies the classification results. Therefore, how the appropriate attribute values can be defined will be very critical for rule derivations by using the RST. In this study, the authors intend to introduce the particle swarm optimization (PSO) algorithm to adjust the attribute setting scopes as an optimization problem to derive the most appropriate attribute values in a complex information system. Finally, the efficiency of the proposed method will be benchmarked with other algorithms by using the Fisher’s iris data set. Based on the benchmark results, the simpler rules can be generated and better classification performance can be achieved by using the PSO based attribute coding method.
KeywordsParticle Swarm Optimization (PSO) Rough Set Theory (RST) Attribute Coding optimization
Unable to display preview. Download preview PDF.
- 1.Aha, D.W., Kibler, D.: Detecting and removing noisy instances from concept de-scriptions, Tecnical Report, University of California, Irvine, 88–12 (1989)Google Scholar
- 2.Chang, J.R., Jheng, Y.H.: Optimization of α-cut Value by using Genetic Algorithm for Fuzzy-based rules extraction. In: The 18th National Conference on Fuzzy and Its Applications, pp. 678–683 (2010)Google Scholar
- 5.Hirsh, H.: Incremental version-space merging: a general framework for concept learning, Ph.D. Thesis, Stanford University (1990)Google Scholar
- 8.Jang, J.R., Hung, J.T.: Rough set theory inference to study pavement maintenance, National Science Council Research Project. Mingshin University of Science and Technology, Hsinchu, Taiwan (2005)Google Scholar
- 9.Jia, G., Zhang, W.: Using PSO to Reliability Analysis of PC Pipe Pile. In: The 3rd International Symposium on Computational Intelligence, vol. 1, pp. 68– 71 (2010)Google Scholar
- 10.Jin, J., Wang, Y., Wang, Q., Yang, B.Q.: The VNP-PSO method for medi-cal image registration.In: The 29th Chinese Control Conference, pp. 5203–5205 (2010)Google Scholar
- 11.Kennedy, J., Eberhart, R.C.: Particle Swarm Optimization. In: Proceeding of the IEEE International Conference on Neural Networks, pp. 12–13 (1995)Google Scholar
- 12.Khamsawang, S., Wannakarn, P., Jiriwibhakorn, S.: Hybrid PSO-DE for solving the economic dispatch problem with generator constraints. In: 2010 2nd International Conference on Computer and Automation Engineering, vol. 5, pp. 135–139 (2010)Google Scholar
- 13.Kim, D.J., Ferrin, D.L., RaghavRao, H.: A study of the effect of consumer trust on consumer expectations and satisfaction: The Korean experience. In: Proceedings of the 5th International Conference on Electronic Commerce, Pittsburgh. ACM International Conference Proceeding Series, pp. 310–315 (2003)Google Scholar
- 14.Kondo, Y., Phimmasone, V., Ono, Y., Miyatake, M.: Verification of efficacy of PSO-based MPPT for photovoltaics. In: 2010 International Conference on Electrical Machines and Systems, pp. 593–596 (2010)Google Scholar
- 17.Quinlan, J.R., Compton, P.J., Horn, K.A., Lazarus, L.A.: Inductive knowledge ac-quisition: A case study. In: Quinlan, J.R. (ed.) Applications of Expert systems, Addison-Wesley, Wokingham (1987)Google Scholar