Relaxation Labeling for Cell-phase Identification

Dat Tran, T Pham

    Research output: A Conference proceeding or a Chapter in BookConference contribution

    Abstract

    Gaussian mixture model (GMM) is used in cell phase identification to model the distribution of cell feature vectors. The model parameters, which are mean vectors, covariance matrices and mixture weights, are trained in an unsupervised learning method using the expectation maximization algorithm. Experiments have shown that the GMM is an effective method capable of achieving high identification rate. However, the GMM approach is not always effective because of ambiguity inherently existing in the cell phase data. To enhance the effectiveness of the GMM for solving this specific problem, the relaxation labeling (RL) is proposed to be used with the GMM. The RL algorithm is a parallel algorithm that updates the probabilities of cell phases by using correlation or mutual information between cell phases to reduce uncertainty among GMMs having overlapping properties.
    Original languageEnglish
    Title of host publicationProceedings: 2006 IEEE International Conference on Systems, Man and Cybernetics
    EditorsBing-Fei Wu, Chung-Cheng Chiu
    Place of PublicationTaiwan
    PublisherIEEE, Institute of Electrical and Electronics Engineers
    Pages1275-1280
    Number of pages6
    ISBN (Print)1424400996
    DOIs
    Publication statusPublished - 2006
    EventIEEE International Conference on Systems, Man, and Cybernetics - Taipei, Taiwan, Province of China
    Duration: 8 Oct 200611 Oct 2006

    Conference

    ConferenceIEEE International Conference on Systems, Man, and Cybernetics
    CountryTaiwan, Province of China
    CityTaipei
    Period8/10/0611/10/06

    Fingerprint

    Labeling
    Identification (control systems)
    Unsupervised learning
    Covariance matrix
    Parallel algorithms
    Experiments

    Cite this

    Tran, D., & Pham, T. (2006). Relaxation Labeling for Cell-phase Identification. In B-F. Wu, & C-C. Chiu (Eds.), Proceedings: 2006 IEEE International Conference on Systems, Man and Cybernetics (pp. 1275-1280). Taiwan: IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/ICSMC.2006.384890
    Tran, Dat ; Pham, T. / Relaxation Labeling for Cell-phase Identification. Proceedings: 2006 IEEE International Conference on Systems, Man and Cybernetics. editor / Bing-Fei Wu ; Chung-Cheng Chiu. Taiwan : IEEE, Institute of Electrical and Electronics Engineers, 2006. pp. 1275-1280
    @inproceedings{837a96ed9bf74fecae0e9276ba8d961c,
    title = "Relaxation Labeling for Cell-phase Identification",
    abstract = "Gaussian mixture model (GMM) is used in cell phase identification to model the distribution of cell feature vectors. The model parameters, which are mean vectors, covariance matrices and mixture weights, are trained in an unsupervised learning method using the expectation maximization algorithm. Experiments have shown that the GMM is an effective method capable of achieving high identification rate. However, the GMM approach is not always effective because of ambiguity inherently existing in the cell phase data. To enhance the effectiveness of the GMM for solving this specific problem, the relaxation labeling (RL) is proposed to be used with the GMM. The RL algorithm is a parallel algorithm that updates the probabilities of cell phases by using correlation or mutual information between cell phases to reduce uncertainty among GMMs having overlapping properties.",
    author = "Dat Tran and T Pham",
    year = "2006",
    doi = "10.1109/ICSMC.2006.384890",
    language = "English",
    isbn = "1424400996",
    pages = "1275--1280",
    editor = "Bing-Fei Wu and Chung-Cheng Chiu",
    booktitle = "Proceedings: 2006 IEEE International Conference on Systems, Man and Cybernetics",
    publisher = "IEEE, Institute of Electrical and Electronics Engineers",
    address = "United States",

    }

    Tran, D & Pham, T 2006, Relaxation Labeling for Cell-phase Identification. in B-F Wu & C-C Chiu (eds), Proceedings: 2006 IEEE International Conference on Systems, Man and Cybernetics. IEEE, Institute of Electrical and Electronics Engineers, Taiwan, pp. 1275-1280, IEEE International Conference on Systems, Man, and Cybernetics, Taipei, Taiwan, Province of China, 8/10/06. https://doi.org/10.1109/ICSMC.2006.384890

    Relaxation Labeling for Cell-phase Identification. / Tran, Dat; Pham, T.

    Proceedings: 2006 IEEE International Conference on Systems, Man and Cybernetics. ed. / Bing-Fei Wu; Chung-Cheng Chiu. Taiwan : IEEE, Institute of Electrical and Electronics Engineers, 2006. p. 1275-1280.

    Research output: A Conference proceeding or a Chapter in BookConference contribution

    TY - GEN

    T1 - Relaxation Labeling for Cell-phase Identification

    AU - Tran, Dat

    AU - Pham, T

    PY - 2006

    Y1 - 2006

    N2 - Gaussian mixture model (GMM) is used in cell phase identification to model the distribution of cell feature vectors. The model parameters, which are mean vectors, covariance matrices and mixture weights, are trained in an unsupervised learning method using the expectation maximization algorithm. Experiments have shown that the GMM is an effective method capable of achieving high identification rate. However, the GMM approach is not always effective because of ambiguity inherently existing in the cell phase data. To enhance the effectiveness of the GMM for solving this specific problem, the relaxation labeling (RL) is proposed to be used with the GMM. The RL algorithm is a parallel algorithm that updates the probabilities of cell phases by using correlation or mutual information between cell phases to reduce uncertainty among GMMs having overlapping properties.

    AB - Gaussian mixture model (GMM) is used in cell phase identification to model the distribution of cell feature vectors. The model parameters, which are mean vectors, covariance matrices and mixture weights, are trained in an unsupervised learning method using the expectation maximization algorithm. Experiments have shown that the GMM is an effective method capable of achieving high identification rate. However, the GMM approach is not always effective because of ambiguity inherently existing in the cell phase data. To enhance the effectiveness of the GMM for solving this specific problem, the relaxation labeling (RL) is proposed to be used with the GMM. The RL algorithm is a parallel algorithm that updates the probabilities of cell phases by using correlation or mutual information between cell phases to reduce uncertainty among GMMs having overlapping properties.

    U2 - 10.1109/ICSMC.2006.384890

    DO - 10.1109/ICSMC.2006.384890

    M3 - Conference contribution

    SN - 1424400996

    SP - 1275

    EP - 1280

    BT - Proceedings: 2006 IEEE International Conference on Systems, Man and Cybernetics

    A2 - Wu, Bing-Fei

    A2 - Chiu, Chung-Cheng

    PB - IEEE, Institute of Electrical and Electronics Engineers

    CY - Taiwan

    ER -

    Tran D, Pham T. Relaxation Labeling for Cell-phase Identification. In Wu B-F, Chiu C-C, editors, Proceedings: 2006 IEEE International Conference on Systems, Man and Cybernetics. Taiwan: IEEE, Institute of Electrical and Electronics Engineers. 2006. p. 1275-1280 https://doi.org/10.1109/ICSMC.2006.384890