Pozyskiwanie wiedzy z danych przy wykorzystaniu klasyfikatorów złożonych

• Autorzy: Ewa Szpunar-Huk

Warianty tytułu

Knowledge Acquisition from Data Using Ensemble Classifiers

• Języki publikacji: PL

Abstrakty

Omówiono podstawy i metody budowy klasyfikatorów złożonych, które to klasyfikatory stanowią obecnie jeden z bardziej dynamicznie rozwijający się kierunek dziedziny pozyskiwania wiedzy z danych. Zaprezentowano wyniki badania, którego celem było sprawdzenie możliwości poprawy jakości pojedynczego drzewa decyzyjnego, wygenerowanego z algorytmem pozyskującym wiedzę systemu klasyfikatorów, poprzez wstępną redukcję komitetu przy użyciu zaproponowanej metody wyboru drzew.

EN

The article discusses main problems connected to the issues of knowledge discovery from data using heterogenous ensemble classifiers (or committee). There are presented main types of such classifiers and their architectures, methods of building and ways of making decisions. In this study are described main achievements in this domain and theoretical backgrounds, which explain principles and interesting properties of committee classifiers. The article also points to the need of knowledge extraction from ensemble of classifiers and within the framework of this domain there is presented a short survey of some used techniques and knowledge extraction methods. The study contains proposal of the method of knowledge extraction from ensemble classifiers based on reduction of number of simple classifiers , which are a part of ensemble (in particularly decision trees), and on using Trepan algorithm. To show the legitimacy of proposed method there are presented descriptions of experiments, along with analyses of their results. As a completion there are presented proposals of some improvements, which demand yet further analysis and research.(original abstract)

Słowa kluczowe

PL

Zarządzanie wiedzą; Systemy z bazą wiedzy; Projektowanie baz danych; Przegląd literatury; Drzewo decyzyjne

EN

Knowledge management; Knowledge based systems; Database design; Literature review; Decision tree

• Czasopismo: Prace Naukowe Akademii Ekonomicznej we Wrocławiu
• Rocznik: 2005
• Numer: nr 1064 Pozyskiwanie wiedzy i zarządzanie wiedzą
• Strony: 268--279

Twórcy

autor

Ewa Szpunar-Huk

Bibliografia

• Ali K.M., Pazzani M.J., Error Reduction Through Learning Multiple Descriptions, "Machine Learning", 1996.
• Breiman L., Bagging Predictors, "Machine Learning", 2000.
• Brown G., Wyatt J., Harris R., Yao Xin, Diversity Creation Methods: A Survey and Categorisation Information Fusion Journal, (Special issue on Diversity in Multiple Classifier Systems), 2004.
• Chan P.K., Stolfo S.J., A Comparative Evaluation of Voting and Meta-learning on Partitioned Data, Proc. Twelfth International Conference on Machine Learning,. Tahoe City, CA: Morgan Kaufmann, 1995.
• Chawla N.V., Moore T.E., Hall L.O., Bowyer K.W., Kegelmeyer W.P., Bagging is a Small--Data-Set Phenomenon, Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, Hawaii, Springer C., 2001.
• Cherkauer K., Human Expert-Level Performance on a Scientific Image Analysis Task by a System Using Combined Artificial Neural Networks, Thirteenth National Conference on Artificial Intelligence. Portland, ÄA AI Press (1996).
• Chu Fang, Zaniolo C., Fast and Light Boosting for Adaptive Mining of Data Streams, PAKDD, 2004.
• Craven M.W., Shavlik J.W., Extracting tree-structured representations of trained networks, [in:] Touretzky D., Mozer M., and Hasselmo M. (eds.), Advances in Neural Information Processing Systems 8, Cambridge, MA: MIT Press, 1996.
• Dietterich T.G., Ensemble methods inmachine learning, [in:] J. Kittler and F. Roli, (eds.), Multiple Classifier Systems, First International Workshop, MCS 2000, Cagliari, Italy, Springer-- Verlag, 2000.
• Estruch V., Ferri C., Hernândez-Orallo J., Ramirez-Quintana M.J., Re-designing Cost-sensitive Decision Tree Learning, [in:] Herrera F., Riquelme J.C., Aguilar J.S., Workshop of "Data Mining and Learning" in the "VIII Conferencia Iberoamericana de Inteligencia Artificial", Ibera-mia'2002, Universidad de Sevilla, pp. 33-42, 2002.
• Ferri C., Hernândez-Orallo J., Ramirez-Quintana M.J., From Ensemble Methods to Comprehensible Models, Proceedings of the 5th International Conference on Discovery Science, 2002.
• Freund Y., lyer R., Schapire R.E., Singer Y., An Efficient Boosting Algorithm for Combining Preferences, ICML-98.
• Freund Y., Schapire R. E., Experiments with a New Boosting Algorithm, in Proceedings of the 13th International Conference on Machine Learning, 1996.
• Kuncheva L, Bezdek J., Sutton M.A., On Combining Multiple Classifiers by Fuzzy Templates, Proceedings of the IEEE Conference of the North American Fuzzy Information Processing Society, Piscataway, NJ: IEEE, 1998.
• Opitz D., Feature Selection for Ensembles, [in:] Proceedings of 16th National Conference on Artificial Intelligence, AAAI, 1999.
• Park B., Kargupta H., Constructing Simpler Decision Trees from Ensemble Models Using Fourier Analysis, Proceedings of the 7th Workshop on Research Issues in Data Mining and Knowledge Discovery, ACM SIGMOD 2002.
• Quinlan J.R., C4.5 Programs for Machine Learning, Morgan Kauffman, 1993.
• Raviv Y., Intrator N., Bootstrapping with Noise: An Effective Régularisation Technique, "Connection Science", 1996.
• Seewald A.K., Petrak J., Widmer G., Hybrid Decision Tree Learners with Alternative Leaf Classifiers: An Empirical Study, FLAIRS-2001, AAAI Press, Menlo Park, California, 2001.
• UCI Machine Learning Repository, [http://www.ics.uci.edu/~mlearn/MLRepository.html].
• Valentini G., Masulli P., Ensembles of Learning Machines, [in:] Neural Nets WIRN Vietri-02, Series Lecture Notes in Computer Sciences, M. Marinaro and R. Tagliaferri, Eds.: Springer-Verlag, Heidelberg (Germany), 2002.
• Wang W., Jones P., Partridge D., Diversity between Neural Networks and Decision Trees for Building Multiple Classier Systems, [in:] Proc. Int. Workshop on Multiple Classier Systems (LNCS 1857), Springer, Calgiari, Italy, 2000.
• [23] Wang H., Fan W., Yu P., Han J., Mining Concept-drifting Data Streams Using Ensemble Classifiers. In ACM SIGKDD, 2003.