Non-logarithmic information measures, α-weighted em algorithms and speedup of learning

Y. Matsuyama

    研究成果: Conference contribution

    3 引用 (Scopus)

    抄録

    Starting from Renyi's α-divergence, a class of generalized EM algorithms called the α-EM algorithms of the WEM algorithms are derived. Merits of this generalization are found on speedup of learning, i.e., acceleration of convergence. Discussions include novel α-versions of logarithm, efficient scores, information matrices and the Cramer-Rao bound. The speedup is examined on Gaussian mixture learning systems.

    元の言語English
    ホスト出版物のタイトルIEEE International Symposium on Information Theory - Proceedings
    ページ385
    ページ数1
    DOI
    出版物ステータスPublished - 1998
    イベント1998 IEEE International Symposium on Information Theory, ISIT 1998 - Cambridge, MA
    継続期間: 1998 8 161998 8 21

    Other

    Other1998 IEEE International Symposium on Information Theory, ISIT 1998
    Cambridge, MA
    期間98/8/1698/8/21

    Fingerprint

    Information Measure
    EM Algorithm
    Speedup
    Acceleration of Convergence
    Cramér-Rao Bound
    Gaussian Mixture
    Information Matrix
    Learning Systems
    Logarithm
    Divergence
    Cramer-Rao bounds
    Learning systems
    Learning
    Generalization
    Class

    ASJC Scopus subject areas

    • Applied Mathematics
    • Modelling and Simulation
    • Theoretical Computer Science
    • Information Systems

    これを引用

    Matsuyama, Y. (1998). Non-logarithmic information measures, α-weighted em algorithms and speedup of learning. : IEEE International Symposium on Information Theory - Proceedings (pp. 385). [708990] https://doi.org/10.1109/ISIT.1998.708990

    Non-logarithmic information measures, α-weighted em algorithms and speedup of learning. / Matsuyama, Y.

    IEEE International Symposium on Information Theory - Proceedings. 1998. p. 385 708990.

    研究成果: Conference contribution

    Matsuyama, Y 1998, Non-logarithmic information measures, α-weighted em algorithms and speedup of learning. : IEEE International Symposium on Information Theory - Proceedings., 708990, pp. 385, 1998 IEEE International Symposium on Information Theory, ISIT 1998, Cambridge, MA, 98/8/16. https://doi.org/10.1109/ISIT.1998.708990
    Matsuyama Y. Non-logarithmic information measures, α-weighted em algorithms and speedup of learning. : IEEE International Symposium on Information Theory - Proceedings. 1998. p. 385. 708990 https://doi.org/10.1109/ISIT.1998.708990
    Matsuyama, Y. / Non-logarithmic information measures, α-weighted em algorithms and speedup of learning. IEEE International Symposium on Information Theory - Proceedings. 1998. pp. 385
    @inproceedings{2d4764f51f0b4615a41da6dc0577b623,
    title = "Non-logarithmic information measures, α-weighted em algorithms and speedup of learning",
    abstract = "Starting from Renyi's α-divergence, a class of generalized EM algorithms called the α-EM algorithms of the WEM algorithms are derived. Merits of this generalization are found on speedup of learning, i.e., acceleration of convergence. Discussions include novel α-versions of logarithm, efficient scores, information matrices and the Cramer-Rao bound. The speedup is examined on Gaussian mixture learning systems.",
    author = "Y. Matsuyama",
    year = "1998",
    doi = "10.1109/ISIT.1998.708990",
    language = "English",
    isbn = "0780350006",
    pages = "385",
    booktitle = "IEEE International Symposium on Information Theory - Proceedings",

    }

    TY - GEN

    T1 - Non-logarithmic information measures, α-weighted em algorithms and speedup of learning

    AU - Matsuyama, Y.

    PY - 1998

    Y1 - 1998

    N2 - Starting from Renyi's α-divergence, a class of generalized EM algorithms called the α-EM algorithms of the WEM algorithms are derived. Merits of this generalization are found on speedup of learning, i.e., acceleration of convergence. Discussions include novel α-versions of logarithm, efficient scores, information matrices and the Cramer-Rao bound. The speedup is examined on Gaussian mixture learning systems.

    AB - Starting from Renyi's α-divergence, a class of generalized EM algorithms called the α-EM algorithms of the WEM algorithms are derived. Merits of this generalization are found on speedup of learning, i.e., acceleration of convergence. Discussions include novel α-versions of logarithm, efficient scores, information matrices and the Cramer-Rao bound. The speedup is examined on Gaussian mixture learning systems.

    UR - http://www.scopus.com/inward/record.url?scp=84871368049&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84871368049&partnerID=8YFLogxK

    U2 - 10.1109/ISIT.1998.708990

    DO - 10.1109/ISIT.1998.708990

    M3 - Conference contribution

    AN - SCOPUS:84871368049

    SN - 0780350006

    SN - 9780780350007

    SP - 385

    BT - IEEE International Symposium on Information Theory - Proceedings

    ER -