Abstract
Machine Learning can be divided into two schools of thought: generative model learning and discriminative model learning. While the MCS community has been focused mainly on the latter, our paper is concerned with questions that arise from ensembles of generative models. Generative models provide us with neat ways of thinking about two interesting learning issues: model selection and semi-supervised learning. Preliminary results show that for semi-supervised low-variance generative models, traditional MCS techniques like Bagging and Random Subspace Method (RSM) do not outperform the single classifier approach. However, RSM introduces diversity between base classifiers. This starting point suggests that diversity between base components has to lie within the structure of the base classifier, and not in the dataset, and it highlights the need for novel generative ensemble learning techniques.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Zhu, X.: Semi-supervised learning literature survey. Technical Report 1530, Computer Sciences, University of Wisconsin-Madison (2005)
Roli, F.: Semi-supervised multiple classifier systems: Background and research directions. In: Oza, N.C., Polikar, R., Kittler, J., Roli, F. (eds.) MCS 2005. LNCS, vol. 3541, pp. 1–11. Springer, Heidelberg (2005)
Blum, A., Mitchell, T.: Combining labeled and unlabeled data with co-training. In: Proc. of the 11th conf. on COLT, pp. 92–100 (1998)
Li, M.: Tri-training: Exploiting unlabeled data using three classifiers. IEEE Trans. on Knowledge and Data Engineering 17(11), 1529–1541 (2005)
Bennett, K.P., Demiriz, A., Maclin, R.: Exploiting unlabeled data in ensemble methods. In: Proc. of the 8th Int. Conf. on KDD, pp. 289–296 (2002)
Leskes, B.: The value of agreement, a new boosting algorithm. In: Auer, P., Meir, R. (eds.) COLT 2005. LNCS, vol. 3559, pp. 95–110. Springer, Heidelberg (2005)
Miller, D., Uyar, S.: A mixture of experts classifier with learning based on both labelled and unlabelled data. In: Proc. of Advances in NIPS, vol. 9, pp. 571–578 (1997)
Buc, F., Grandvalet, Y., Ambroise, C.: Semi-supervised marginboost. In: Proc. of Advances in NIPS, vol. 14 (2002)
Ng, A., Jordan, M.: On generative vs. discriminative classifiers: A comparison of logistic regression and naive bayes. In: Proc. of Advances in NIPS, vol. 15 (2002)
Cozman, F.G., Cohen, I., Cirelo, M.C., et al.: Semi-supervised learning of mixture models. In: 20th International Conference on Machine Learning, pp. 99–106 (2003)
MartÃnez, C., Fuentes, O.: Face Recognition Using Unlabeled Data. Computación y Sistemas 7(2), 123–129
Balcan, M., Blum, A.: An Augmented PAC Model for Semi-Supervised Learning. In: Chapelle, O., et al. (eds.) Semi-Supervised Learning. The MIT Press, Cambridge (2006)
Dietterich, T.G.: Ensemble methods in machine learning. In: Kittler, J., Roli, F. (eds.) MCS 2000. LNCS, vol. 1857, pp. 1–15. Springer, Heidelberg (2000)
Kuncheva, L.: Combining Pattern Classifiers: Methods and Algorithms. Wiley Press, Chichester (2004)
Nigam, K., McCallum, A.K., Thrun, S., Mitchell, T.M.: Text classification from labeled and unlabeled documents using EM. Machine Learning 39(2/3) (2000)
Yang, Y., Webb, G., et al.: To select or to weigh: A comparative study of model selection and model weighing for spode ensembles. In: Fürnkranz, J., Scheffer, T., Spiliopoulou, M. (eds.) ECML 2006. LNCS, vol. 4212, pp. 533–544. Springer, Heidelberg (2006)
Brown, G.: A new perspective on information theoretic feature ranking. In: 12th Int. Conf. on Artificial Intelligence and Statistics (2009) (to appear)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2009 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Zanda, M., Brown, G. (2009). A Study of Semi-supervised Generative Ensembles. In: Benediktsson, J.A., Kittler, J., Roli, F. (eds) Multiple Classifier Systems. MCS 2009. Lecture Notes in Computer Science, vol 5519. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-02326-2_25
Download citation
DOI: https://doi.org/10.1007/978-3-642-02326-2_25
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-02325-5
Online ISBN: 978-3-642-02326-2
eBook Packages: Computer ScienceComputer Science (R0)