三参数Weibull分布有限混合模型及应用

详细信息    查看全文 | 推荐本文 |

英文篇名：Three-parameter Weibull Distribution FMM and Its Application 作者：刘福香 英文作者：Liu Fuxiang;School of Economics, Harbin University of Commerce; 关键词：有限混合模型 ; 三参数Weibull分布 ; 分组数据 英文关键词：FFM;;three-parameter Weibull distribution;;grouped data 中文刊名：TJJC 英文刊名：Statistics & Decision 机构：哈尔滨商业大学经济学院; 出版日期：2018-08-17 14:59 出版单位：统计与决策 年：2018 期：v.34;No.507 基金：国家自然科学基金资助项目(31670476);; 国家社会科学基金资助项目(17BJY119);; 黑龙江省哲学社会科学项目(16JYE01);; 哈尔滨商业大学博士启动基金项目(14rw12);哈尔滨商业大学博士后资助项目(2017BSH029) 语种：中文; 页：TJJC201815004 页数：4 CN：15 ISSN：42-1009/C 分类号：16-19

摘要

为展示有限混合模型的原理及应用,文章利用调查的大兴安岭地区混交林直径分布这类频数数据或分组数据,采用三参数Weibull分布的混合,拟合混交林直径的分布特征。并与传统方法进行比较,结果表明,FMM拟合效果优于传统模型,特别是对于数据是多峰、有偏和结尾等数据类型。
        In order to demonstrate the principle and application of Finite Mixture Model(FMM), this paper uses such frequency data or grouped data as diameter distribution of mixed-species forest stands of Daxing'anlin region to fit the characteristics of mixed-species forest stands by adopting the mixture of three-parameter Weibull distribution. And finally the paper also makes a comparison with the traditional methods. The study result shows that the fitting effect of the FMM is better than that of the traditional model, especially for the data types of multiple peak, bias and end tail.

引文

[1]Macdnald P D M,Pitcher T J.Age Groups From Size Frequency Da-ta:A Versatile and Efficient Method of Analyzing Distribution Mix-tures[J].J.Fish.Res.Board Can,1979.
    [2]Zasada M Cieszewski C J.A Finite Mixture Distribution Approach forCharacterizing Tree Diameter Distributions by Natural Social Class inPure Even-aged Scots Pine Stand in Poland[J].For.Eco.Manag,2005.
    [3]Zhang L,Liu C.Fitting Irregular Diameter Distributions of ForestStands by Weibull,Modified Weibull,and Mixture Weibull Models[J].J.For.Res,2006.
    [4]Chandra,S.On the Mixtures of Probability Distributions.Scand[J].J.Stat,1977,(4).
    [5]Everitt B S,Hand D J.Finite Mixture Distributions[M].Cambridge:Cambridge University Press,1981.
    [6]Titterington D M,Smith A F M,Makov U E.Statistical Analysis of Fi-nite Mixture Distributions[M].New York:John Wiley&Sons,1985.
    [7]Bartholomew D J.Sufficient Conditions for a Mixture of Exponentialsto be a Probability Density Function[J].Ann.Math.Stat,1969,(40).