Estimation of the Variance for the Maximum Likelihood Estimates in Normal Mixture Models and Normal Hidden Markov Models

• IQBAL Muhammad

  Graduate School of Science and Engineering, Saga University

• NISHI Akihiro

  Faculty of Culture and Education, Saga University

• KIKUCHI Yasuki

  Graduate School of Biomedical Sciences, Nagasaki University

• NOMAKUCHI Kentaro

  Faculty of Science, Kochi University

抄録

In this article, we derive the observed information matrices for normal mixture models and normal hidden Markov models. We also describe the parametric bootstrap method for the said models. The matrices and the method mentioned above are used to estimate the variance of the maximum likelihood estimates (MLEs) obtained by the Expectation-Maximization (EM) algorithm. Finally, a numerical example is shown using a data set named " faithful " given in the free statistical software R.

収録刊行物

• Ｊｏｕｒｎａｌ　ｏｆ　ｔｈｅ　Ｊａｐａｎｅｓｅ　Ｓｏｃｉｅｔｙ　ｏｆ　Ｃｏｍｐｕｔａｔｉｏｎａｌ　Ｓｔａｔｉｓｔｉｃｓ

  Ｊｏｕｒｎａｌ　ｏｆ　ｔｈｅ　Ｊａｐａｎｅｓｅ　Ｓｏｃｉｅｔｙ　ｏｆ　Ｃｏｍｐｕｔａｔｉｏｎａｌ　Ｓｔａｔｉｓｔｉｃｓ 24 (1), 39-66, 2011

  日本計算機統計学会

キーワード

• EM algorithm
• Observed information matrix
• Parametric bootstrap

詳細情報

• CRID: 1390282679390985216
• NII論文ID: 10030497169
• NII書誌ID: AA10823693
• DOI: 10.5183/jjscs.1002001_183
• ISSN: 18811337
• ISSN: 09152350
• NDL書誌ID: 030632929
• Web Site: https://doi.org/10.5183
• Web Site: https://ndlsearch.ndl.go.jp/books/R000000004-I030632929
• Web Site: https://www.jstage.jst.go.jp/article/jjscs/24/1/24_jjscs.1002001_183/_pdf
• 本文言語コード: en
• データソース種別:
  • JaLC
  • NDL
  • Crossref
  • CiNii Articles
• 抄録ライセンスフラグ: 使用不可