Badanie związków między miarami zróżnicowania błędów pojedynczych modeli a jakością modelu zagregowanego

• Autorzy: Dorota Rozmus

Warianty tytułu

Analysis of Relationship Between Diversity Measures and Error of Aggregated Classifier

• Języki publikacji: PL

Abstrakty

W dalszej części artykułu skupiono się jedynie na miarach globalnych. Wszystkie one wymagają uprzedniego przekodowania wyników klasyfikacji na wartość 0, gdy na podstawie modelu uzyskano poprawną klasyfikację, i na wartość 1 w przypadku przeciwnym. (fragment tekstu)

EN

The main aim of ensembles is improvement of their classification and prediction accuracy. One of the elements required for accurate prediction when using an ensemble is recognised to be error "diversity".
The paper presents different diversity measures and discuss their main properties. It shows also the influence of different factors (eg. number of single models, the training set size) on the diversity of base classifiers of the ensemble. (original abstract)

Słowa kluczowe

PL

Miary zróżnicowania

EN

Diversity measures

• Czasopismo: Prace Naukowe Akademii Ekonomicznej we Wrocławiu. Taksonomia
• Rocznik: 2006
• Tom: 13
• Numer: nr 1126 Klasyfikacja i analiza danych - teoria i zastosowania
• Strony: 214--223

Twórcy

autor

Dorota Rozmus

• Akademia Ekonomiczna im. Karola Adamieckiego w Katowicach

Bibliografia

• Blake C., Keogh E., Merz C.J. (1998), UCI Repository of Machine Learning Databases, Department of Information and Computer Science, University of California, Irvine.
• Cunningham P., Carney J. (2000), Diversity Versus Quality in Classification Ensembles Based on Feature Selection, Technical Report TCD-CS-2000-02, Department of Computer Science, Trinity College, Dublin.
• Fleiss J.L. (1981), Statistical Methods for Rates and Proportions, John Willey and Sons.
• Hansen L.K., Salamon P. (1990), Neural Networks Ensembles, "IEEE Transactions on Pattern Analysis and Machine Intelligence" 12 (10).
• Kohavi R., Wolpert D.H. (1996), Bias Plus Variance Decomposition for Zero-one Loss Function, "Proceedings of the 12th International Conference on Machine Learning", Morgan Kaufmann, Tahoe City.
• Kuncheva L.I., Whitaker C.J. (2001), Feature Subsets for Classifier Combination: an Enumerative Experiment, "Proceedings of the 2nd International Workshop on Multiple Classifier Systems", Springer-Verlag, Cambridge, UK.
• Kuncheva L.I., Whitaker C.J. (2003), Measures of Diversity in Classifier Ensembles, "Machine Learning" 51.
• Krogh A., Vedelsby J. (1995), Neural Network Ensembles, Cross Validation, and Active Learning, NIPS 7.
• Opitz D., Shavlik J. (1996), Generating Accurate and Diverse Members of a Neural Network Ensemble, "Advances in Neural Information Processing Systems", МIT Press, Denver.
• Partridge D., Krzanowski W. (1997), Software Diversity: Practical Statistics for Its Measurement and Exploitation, "Information and Software Technology" 39.
• Ruta D., Gabrys В. (2002a), A Theoretical Analysis of the Limits of Majority Voting Errors for Multiple Classifier Systems, "Pattern Analysis and Applications" 5(4).
• Ruta D., Gabrys В. (2002b), Set Analysis of Coincident Errors and Its Applications for Combining Classifiers, "Pattern Recognition and String Maching, Combinatorial Optimisation", 13, Kluwer Academic Publishers.
• Zenobi G., Cunningham P. (2001), Using Diversity in Preparing Ensembles of Classifiers Based on Different Feature Subsets to Minimise Generalisation Error, "Proceedings of the 12th European Conference on Machine Learning".