Publications

Under review

  • Consistency and generalization bounds for maximum entropy Estimation
    S. Wang,  R. Greiner and S. Wang. [abstract] [Long Version][Short Version]
  • Exploiting syntactic, semantic and lexical regularities in language modeling via directed Markov random fields
    S. Wang, S. Wang R. Greiner, D. Schuurmans and L. Cheng. [pdf]
  • The latent maximum entropy principle
    S. Wang, D. Schuurmans and Y. Zhao. [pdf]

2008

  • Boosting with incomplete information
    G. Haffari, Y. Wang, S. Wang, G. Mori and F. Jiao
    The 25th International Conference on Machine Learning, ICML-2008. [pdf]
  • Segmenting brain tumors using pseudo--conditional random fields
    C. Lee, S. Wang, A. Murtha, M. Brown and R. Greiner
    The 11th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI-2008. [pdf]
  • Constrained classification on structured data
    C. Lee, M. Brown, R. Greiner, S. Wang and A. Murtha
    The 23th AAAI Conference on Artificial Intelligence, AAAI-2008.
  • Unsupervised discovery of compound entities for relationship extraction
    C. Ramakrishnan, P. Mendes, S. Wang and A. Sheth
    The 16th International Conference on Knowledge Engineering and Knowledge Management Knowledge Patterns, EKAW-2008.

2006

  • Almost sure convergence of Titterington's recursive estimator for finite mixture
    S. Wang and Y. Zhao
    Statistics & Probability Letters, Vol. 76, No. 18, pp. 2001-2006, December 2006. [pdf]
  • Learning to model spatial dependency: Semi-supervised discriminative random fields
    C. Lee, S. Wang, F. Jiao, D. Schuurmans and R. Greiner
    Advances in Neural Information Processing Systems, NIPS-2006. [pdf]
  • Implicit online learning with kernels
    L. Cheng, S. Vishwanathan, D. Schuurmans, S. Wang and T. Caelli
    Advances in Neural Information Processing Systems, NIPS-2006. [pdf]
  • Stochastic analysis of lexical and semantic enhanced structural language model
    S. Wang, S. Wang, L. Cheng, R. Greiner and D. Schuurmans
    The 8th International Colloquium on Grammatical Inference, ICGI-2006. [pdf]
  • Semi-supervised conditional random fields for improved sequence segmentation and labeling
    F. Jiao, S. Wang, C. Lee, R. Greiner and D. Schuurmans
    The Joint 21st International Conference on Computational Linguistics and 44th Annual Meeting of the Association for Computational Linguistics, COLING/ACL-2006. [pdf]
  • Using query-specific variance estimates to combine Bayesian classifiers
    C. Lee, R. Greiner and S. Wang
    The 23th International Conference on Machine Learning, ICML-2006. [pdf]
  • An online discriminative approach to background subtraction
    L. Cheng, S. Wang, D. Schuurmans, T. Caelli and S. Vishwanathan
    IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS-2006. [pdf]

2005

  • Consistency and generalization bounds for maximum entropy density estimation
    S. Wang, R. Greiner and S. Wang
    The 16th International Conference on Algorithmic Learning Theory, ALT-2005.
  • Exploiting syntactic, semantic and lexical regularities in language modeling via directed Markov random fields
    S. Wang, S. Wang, R. Greiner, D. Schuurmans and L. Cheng
    The 22th International Conference on Machine Learning, ICML-2005. [pdf]
  • Variational Bayesian image modelling
    L. Cheng, F. Jiao, D. Schuurmans and S. Wang
    The 22th International Conference on Machine Learning, ICML-2005. [pdf]
  • Combining statistical language models via the latent maximum entropy principle
    S. Wang, D. Schuurmans, F. Peng and Y. Zhao
    Machine Learning Journal: Special Issue on Learning in Speech and Language Technologies, Vol. 60, pp. 229-250, 2005. [pdf]

2004

  • Learning mixture models with the regularized latent maximum entropy principle
    S. Wang, D. Schuurmans, F. Peng and Y. Zhao
    IEEE Trans. on Neural Networks: Special Issue on Information Theoretic Learning, Vol. 15, No. 4, pp. 903-916, 2004. [pdf]
  • Augmenting naive Bayes text classifier using statistical n-gram language modeling
    F. Peng, D. Schuurmans and S. Wang
    Information Retrieval, Vol. 7, No. 3-4, pp. 317-345, 2004. [pdf]

2003

  • Learning continuous latent variable models with Bregman divergence
    S. Wang and D. Schuurmans
    The 14th International Conference on Algorithmic Learning Theory, ALT-2003. [pdf]
  • Boltzmann machine learning with the latent maximum entropy principle
    S. Wang, D. Schuurmans, F. Peng and Y. Zhao
    The Nineteenth Conference on Uncertainty in Artificial Intelligence, UAI-2003. [ps]
  • Learning mixture models with the latent maximum entropy principle
    S. Wang, D. Schuurmans, F. Peng and Y. Zhao
    The 20th International Conference on Machine Learning, ICML-2003. [ps]
  • Semantic n-gram language modeling with the latent maximum entropy principle
    S. Wang, D. Schuurmans, F. Peng and Y. Zhao
    International Conference on Acoustics, Speech, ans Signal Processing, ICASSP-2003. [ps]
  • Language and task independent text categorization with simple language models
    F. Peng, D. Schuurmans, S. Wang
    North American Chapter of the Association for Computational Linguistics, NAACL-2003.
  • Language independent authorship attribution with character level n-Gram language modeling
    F. Peng, D. Schuurmans, S. Wang
    The 10th Conference of the European Chapter of the Association for Computational Linguistics, EACL-2003.

2002

  • The latent maximum entropy principle
    S. Wang, R. Rosenfeld, Y. Zhao and D. Schuurmans
    IEEE International Symposium on Information Theory, ISIT-2002. Abstract[ps], [ps]
  • Almost sure convergence of Titterington's recursive estimator for finite mixture models
    S. Wang and Y. Zhao
    IEEE International Symposium on Information Theory, ISIT-2002.
  • Predicting oral reading miscues
    J. Mostow, J. Beck, V. Winter, S. Wang and B. Tobin
    International Conference on Spoken Language Processing, ICSLP-2002.

2001

  • On-line Bayesian tree-structured transformation of HMMs with optimal model selection for speaker adaptation
    S. Wang and Y. Zhao
    IEEE Trans. on Speech and Audio Processing, Vol. 9, No. 6, pp. 663-677, September 2001. [pdf]
  • Latent maximum entropy principle for statistical language modeling
    S. Wang, R. Rosenfeld and Y. Zhao
    IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU-2001.
  • Recursive estimation of time-varying environments for robust speech recognition
    Y. Zhao, S. Wang and K. Yen
    IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP-2001.

2000

  • Optimal on-line Bayesian model selection for speaker adaptation
    S. Wang and Y. Zhao
    International Conference on Spoken Language Processing, ICSLP-2000.
  • On-line Bayesian speaker adaptation by using tree-structured transformation and robust priors
    S. Wang and Y. Zhao
    IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP-2000.

1999

  • On-line Bayesian tree-structured transformation of hidden Markov models for speaker adaptation
    S. Wang and Y. Zhao
    IEEE Workshop on Automatic Speech Recognition and Understanding, ASRU-1999.
  • A unified framework for recursive maximum likelihood estimation of hidden Markov models
    S. Wang and Y. Zhao
    The 33rd Annual Conference on Information Sciences and Systems, CISS-1999.

1998

  • On convergence of maximum likelihood estimation of binary HMMs by EM algorithm
    M. Li, S. Wang and Y. Zhao
    The 32rd Annual Conference on Information Sciences and Systems, CISS-1998, pp. 1018-1024.

1995

  • Short-term generation scheduling with transmission and environmental constraints using an augmented Lagrangian relaxation
    S. Wang, S. Shahidehpour, D. Kirschen, S. Mokhtari and G. Irisarri
    IEEE Trans. on Power Systems, Vol. 10, No. 3, pp. 1294-1301, August 1995. [pdf]
  • Probabilistic marginal cost curve and its applications
    S. Wang, S. Shahidehpour and N. Xiang
    IEEE Trans. on Power Systems, Vol. 10, No. 3, pp 1321-1328, August 1995. [pdf]

Manuscript

  • On determination of domains of convergence for the EM algorithm
    S. Wang and Y. Zhao
    [ps]