ABSTRACT
We consider a setting for discriminative semi-supervised learning where unlabeled data are used with a generative model to learn effective feature representations for discriminative training. Within this framework, we revisit the two-view feature generation model of co-training and prove that the optimum predictor can be expressed as a linear combination of a few features constructed from unlabeled data. From this analysis, we derive methods that employ two views but are very different from co-training. Experiments show that our approach is more robust than co-training and EM, under various data generation conditions.
- Ando, R. K., & Zhang, T. (2005). A framework for learning predictive structures from multiple tasks and unlabeled data. Journal of Machine Learning Research, 6, 1817--1853. Google ScholarDigital Library
- Blum, A., & Mitchell, T. (1998). Combining labeled and unlabeled data with co-training. Proceedings of the eleventh annual conference on Computational learning theory (pp. 92--100). Google ScholarDigital Library
- Dasgupta, S., Littman, M., & McAllester, D. (2001). PAC generalization bounds for co-training. NIPS' 01.Google Scholar
- Nigam, K., McCallum, A. K., Thrun, S., & Mitchell, T. (2000). Text classification from labeled and unlabeled documents using EM. Machine Learning, Special issue on information retrieval, 103--134. Google ScholarDigital Library
- Vapnik, V. (1998). Statistical learning theory. New York: John Wiley & Sons.Google Scholar
- Zhang, T., & Oles, F. J. (2000). A probability analysis on the value of unlabeled data for classification problems. ICML 2000 (pp. 1191--1198).Google Scholar
- Zhu, X., Ghahramani, Z., & Lafferty, J. (2003). Semi-supervised learning using gaussian fields and harmonic functions. ICML 2003.Google ScholarDigital Library
Recommendations
Automatic feature generation for machine learning--based optimising compilation
Recent work has shown that machine learning can automate and in some cases outperform handcrafted compiler optimisations. Central to such an approach is that machine learning techniques typically rely upon summaries or features of the program. The ...
Semi-supervised learning using randomized mincuts
ICML '04: Proceedings of the twenty-first international conference on Machine learningIn many application domains there is a large amount of unlabeled data but only a very limited amount of labeled training data. One general approach that has been explored for utilizing this unlabeled data is to construct a graph on all the data points ...
Combining labeled and unlabeled data with co-training
COLT' 98: Proceedings of the eleventh annual conference on Computational learning theory
Comments