Semi-Supervised Learning by Gaussian Mixtures

정규 혼합분포를 이용한 준지도 학습

  • Published : 2008.10.31


Discriminant analysis based on Gaussian mixture models, an useful tool for multi-class classifications, can be extended to semi-supervised learning. We consider a model selection problem for a Gaussian mixture model in semi-supervised learning. More specifically, we adopt Bayesian information criterion to determine the number of subclasses in the mixture model. Through simulations, we illustrate the usefulness of the criterion.


  1. Breiman, L., Fredman, J., Olshen, R. and Stone, C. (1984). Classification and Regression Trees, Wadsworth, Belmont
  2. Dempster, A. P., Laird, N. M. and Rubin, D. B. (1977). Maximum likelihood from incomplete data via the EM algorithm, Jornal of the Royal Statistical Society, Series B, 39, 1-38
  3. Halbe, Z. and Aladjem, M. (2005). Model-based mixture discriminant analysis|an experimental study, Pattern Recognition, 38, 437-440
  4. Hastie, T. and Tibshirani, R. (1996). Discriminant analysis by Gaussian Mixtures, Journal of the Royal Statististical Society, Serise B, 58, 158-176
  5. Nigam, K., McCallum, A. K., Thrun, S. and Mitchell, T. (2000). Text classification from labeled and unlabeled documents using EM, Machine Learning, 39, 103-134
  6. Zhu X. (2005). Semi-Supervised Learning Literature Survey, Technical Report 1530, Computer Sciences, University of Wisconsin-Madison