The pipe of samples visualization method as base for evolutionary data discretisation

• Autorzy: Czerniak J. , Dobrosielski W.
• Języki publikacji: EN

Abstrakty

EN

This article shows the application of the authors’ own method for visualizing multidimensionality, i.e. so called Pipe of Samples, which makes possible to visualize up to 360 dimensions. This approach constituted the base for development of evolutionary discretisation algorithm dedicated for pre-processing of data to be processed using rough sets theory. The study presents operators of crossing, mutation and selection. Structures of the algorithm data have been prepared on the basis of the aforementioned visualization so that each of the achieved individuals described one complete discretisation solution. Hence, in the proposed approach, the population is a set of many complete discretisations of all the attributes. The solution is optimized by means of evolutionary search for the optimum. The study includes results of experiments that compared LDGen adaptation algorithm with other discretisation methods used in rough sets theory. As main components of the article may be regarded such elements like visualisation method, evolutionary data discretisation method including dedicated operators and discussion on the results of experiments.

Słowa kluczowe

EN

Pipe of Samples; LDGen; discretisation; rough sets theory; genetic algorithm

PL

dyskretyzacja; teoria zbiorów przybliżonych; algorytm interacyjny

• Wydawca: Komisja Informatyki Polskiej Akademii Nauk, Oddział w Gdańsku
• Czasopismo: Metody Informatyki Stosowanej
• Rocznik: 2010
• Tom: nr 3 (23)
• Strony: 57--67
• Opis fizyczny: Bibliogr. 25 poz., rys.

Twórcy

autor

Czerniak J.

autor

Dobrosielski W.

• Casimir the Great University in Bydgoszcz, Institute of Technology, Systems Research Institute, Polish Academy of Sciences

Bibliografia

• [1] Brijs T., Vanhoof K.: Cost-sensitive Discretization of Numeric Attributes, Proceedings of the Second European Symposium on Principles of Data Mining and Knowledge Discovery table of contents, Lecture Notes In Computer Science; Vol. 1510, pp 102-110, 1998.
• [2] Cerquides J., Mantaras R.L.: Proposal and empirical comparison of a parallelizable distance-based discretization method. In: Third International Conference on Knowledge Discovery and Data Mining, pp.139-142, 1997
• [3] Chmielewski M. R., Grzymala-Busse J. W.: Global discretization of continuous attributes as preprocessing for machine learning, [In:] Lin T. Y., Wildberger A. (ed.), Soft Computing: Rough Sets, Fuzzy Logic, Neural Networks, Uncertainty Management, Knowledge Discovery, San Diego, Simulation Councils Inc.,pp 294-297, 1995.
• [4] Cios K. J., Pedrycz W., Swiniarski R. W.: Data mining methods for knowledge discovery, Dordrecht, Kluwer Academic Publishers, 1999
• [5] Czerniak J., Zarzycki H.: Application of rough sets in the presumptive diagnosis of urinary system diseases, in: Artificial Intelligence and Security in Computing Systems, Kluwer Academic Publishers, pp 41-51, 2002.
• [6] Czerniak J., Evolutionary Approach to Data Discretization for Rough Sets Theory, Fundamenta Informaticae, Volume 92 , Issue 1-2 (January 2009), pp.43-61
• [7] Czerniak J., The ’Pipe of Samples’ approach as a method to visualization of multidimensionality – general conception, (in Polish), in: Proc. 7th Symposium of Computer Science, Faculty of Computer Science and Information Systems, Szczecin University of Technology, Informa Press, 2002, Vol. II, pp. 375-381, 2002.
• [9] Doherty P., Łukaszewicz W., Skowron A., Szałas A.: Knowledge representation techniques : a rough set approach, in: Studies in Fuzziness and Soft Computing, Vol. 202, Springer, 2006
• [10] Dougherty J., Kohavi R., Sahami M.: Supervised and unsupervised discretizations of continuous features, in: Proc. 12th Int. Conf. on Machine Learning, Morgan Kaufmann, pp.194-202, 1995.
• [11] Everitt B., Landau S., Leese M.: Cluster Analysis, 4th Edition, Arnold Publishers, London, 2001
• [12] Fayyad U., Irani K.: Multi-interval discretization of continuous-valued attributes for classification learning, in: Proc. Thirteenth International Joint Conference on Artificial Intelligence, San Mateo, CA: Morgan Kaufmann, pp.1022-1027, 1993.
• [13] Grzymala-Busse J. W., Stefanowski J.: Three approaches to numerical attribute discretization for rule induction, in: International Journal of Intelligent Systems, vol. 16 no. 1, pp. 29-38, 2001.
• [14] Holte R. C.: Very simple classification rules perform well on most commonly used datasets, in: Machine Learning, pp. 63-90, 1993.
• [15] Kerber R.: Chimerge: Discretization of numeric attributes, in: Proc. AAAI-92, Ninth National Confrerence Articial Intelligence, AAAI Press/The MIT Press, pp.123-128, 1992.
• [16] Kohavi R., Sahami M.: Error-based and entropy-based discretization of continuos features, in: Proc. of the 2nd Int. Conf. on Knowledge Discovery and Data Mining, Portland, pp. 114-119, 1996
• [17] Nguyen H. S., Skowron A.: Quantization of Real Values Attributes, Rough set and Boolean Reasoning Approaches, in: Proc. of the Second Joint Conference on Information Sciences, Wrightsville Beach, NC, pp. 34-37, 1995.
• [18] Nguyen H. S.: Discretization of real value attributes. Boolean reasoning approach. Ph.D. Thesis, University of Warsaw, Warszawa, 1997
• [19] Polkowski L.: Rough Sets, Mathematical Foundations, Physica-Verlag, Heidelberg, 2002.
• [20] Rakus-Andersson, E.: Fuzzy and rough techniques in medical diagnosis and medication, in: Studies in Fuzziness and Soft Computing, Vol 212, Springer, 2007
• [21] Stefanowski J.: Algorithms of rule induction for knowledge discovery. (In Polish), Habilitation Thesis published as Series Rozprawy no. 361, Poznan Univeristy of Technology Press, Poznan, 2001
• [22] Stefanowski J., Nowaczyk S.: An Experimental Study of Using Rule Induction Algorithm in Combiner Multiple Classifier, in: International Journal of Computational Intelligence Research, Vol.3, No.4, pp. 335-342, 2007
• [23] Susmaga R.: Analyzing discretization of continuous attributes given a monotonic discrimination function, in: Journal Intelligent Data Analysis, Vol. 1, No. 3, pp.157-179, 1997
• [24] Wróblewski J.: Finding minimal reducts using genetic algorithm, in: Proc. of Second Joint Annual Conference on Information Sciences, Wrightsville Beach, North Carolina, 28 September-1 October, USA, pp.186-189, 1995
• [25] Yanushkevich S., Shmerko V., Lu D. C., Adams K., McGregor J.: Spectra of Boolean Functions: Computation of Reed-Muller, in: Arithmetic and Walsh Spectrum via Taylor Expansion, IEEE Trans. Computers, 2004