Semiparametric mixture: Continuous scale mixture approach

In this article, we propose a new estimation procedure for a class of semiparametric mixture models that is a mixture of unknown location-shifted symmetric distributions. The proposed method assumes that the nonparametric symmetric distribution falls in a rich class of continuous normal scale mixture distributions. With this new modeling approach, we can suitably avoid the misspecification problem in traditional parametric mixture models. In addition, unlike some existing semiparametric methods, the proposed method does not require any modification or smoothing of the likelihood as it can directly estimate parametric and nonparametric components simultaneously in the model. Furthermore, the proposed parameter estimates are robust against outliers. The estimation algorithms are introduced and numerical studies are conducted to examine the finite sample performance of the proposed procedure and to compare it with other existing methods.

[1]  Byungtae Seo,et al.  Adaptive robust regression with continuous Gaussian scale mixture errors , 2017 .

[2]  B. Lindsay.  Mixture models : theory, geometry, and applications , 1995 .

[3]  Bradley Efron,et al.  HOW BROAD IS THE CLASS OF NORMAL SCALE MIXTURES , 1978 .

[4]  H. Wynn.  The Sequential Generation of $D$-Optimum Experimental Designs , 1970 .

[5]  Chauveau,et al.  An EM algorithm for a semiparametric mixture model , 2006 .

[6]  M. West.  On scale mixtures of normal distributions , 1987 .

[7]  M. Stephens.  Dealing with label switching in mixture models , 2000 .

[8]  Yong Wang.  On fast computation of the non‐parametric maximum likelihood estimate of a mixing distribution , 2007 .

[9]  Sylvia Frühwirth-Schnatter,et al.  Finite Mixture and Markov Switching Models , 2006 .

[10]  Corwin L. Atwood,et al.  Convergent Design Sequences, for Sufficiently Regular Optimality Criteria , 1976 .

[11]  B. Lindsay.  The Geometry of Mixture Likelihoods: A General Theory , 1983 .

[12]  Douglas Kelker,et al.  Infinite Divisibility and Variance Mixtures of the Normal Distribution , 1971 .

[13]  Akimichi Takemura,et al.  Strong consistency of the maximum likelihood estimator for finite mixtures of location-scale distributions when the scale parameters are exponentially small , 2006 .

[14]  Laurent Bordes,et al.  A stochastic EM algorithm for a semiparametric mixture model , 2007, Comput. Stat. Data Anal..

[15]  Roger W. Johnson,et al.  Exploring Relationships in Body Dimensions , 2003 .

[16]  B. Lindsay.  The Geometry of Mixture Likelihoods, Part II: The Exponential Family , 1983 .

[17]  C. Robert,et al.  Computational and Inferential Difficulties with Mixture Posterior Distributions , 2000 .

[18]  D. Hunter,et al.  Inference for mixtures of symmetric distributions , 2007, 0708.0499.

[19]  B. Lindsay,et al.  Bayesian Mixture Labeling by Highest Posterior Density , 2009 .

[20]  Dankmar Böhning,et al.  A note on the maximum deviation of the scale-contaminated normal to the best normal distribution , 2002 .

[21]  W. Yao,et al.  Minimum profile Hellinger distance estimation for a semiparametric mixture model , 2014 .

[22]  Dankmar Bohning.  Convergence of Simar's Algorithm for Finding the Maximum Likelihood Estimate of a Compound Poisson Process , 1982 .

[23]  Weixin Yao,et al.  Label switching and its solutions for frequentist mixture models , 2015 .

[24]  L. Bordes,et al.  SEMIPARAMETRIC ESTIMATION OF A TWO-COMPONENT MIXTURE MODEL , 2006, math/0607812.

[25]  Dankmar Böhning,et al.  Computer-Assisted Analysis of Mixtures and Applications , 2000, Technometrics.

[26]  David R. Hunter,et al.  An EM-Like Algorithm for Semi- and Nonparametric Estimation in Multivariate Mixtures , 2009 .

[27]  S. Basu.  Existence of a normal scale mixture with a given variance and a percentile , 1996 .

[28]  J. Kiefer,et al.  CONSISTENCY OF THE MAXIMUM LIKELIHOOD ESTIMATOR IN THE PRESENCE OF INFINITELY MANY INCIDENTAL PARAMETERS , 1956 .

[29]  Byungtae Seo,et al.  A new algorithm for maximum likelihood estimation in normal scale-mixture generalized autoregressive conditional heteroskedastic models , 2015 .

[30]  N. Laird.  Nonparametric Maximum Likelihood Estimation of a Mixing Distribution , 1978 .

[31]  J. Kalbfleisch,et al.  An Algorithm for Computing the Nonparametric MLE of a Mixing Distribution , 1992 .

[32]  Yong Wang,et al.  Estimation of finite mixtures with symmetric components , 2011, Statistics and Computing.

[33]  Weixin Yao,et al.  A profile likelihood method for normal mixture with unequal variance , 2010 .

[34]  Changbao Wu,et al.  Some Algorithmic Aspects of the Theory of Optimal Designs , 1978 .

[35]  W. Yao,et al.  Flexible estimation of a semiparametric two-component mixture model with one parametric component , 2015 .

[36]  Daeyoung Kim,et al.  Root selection in normal mixture models , 2012, Comput. Stat. Data Anal..

[37]  H. Wynn.  Results in the Theory and Construction of D‐Optimum Experimental Designs , 1972 .

[38]  Dankmar Böhning,et al.  A vertex-exchange-method in D-optimal design theory , 1986 .

[39]  Geoffrey J. McLachlan,et al.  Finite Mixture Models , 2019, Annual Review of Statistics and Its Application.

[40]  R. Hathaway.  A Constrained Formulation of Maximum-Likelihood Estimation for Normal Mixture Distributions , 1985 .