Extension of Comparative Analysis of Estimation Methods for Dirichlet Distribution Parameters
dirichlet distribution, parameter estimation, maximum likelihood estimator
Abstract
The Dirichlet distribution is a generalization of the Beta distribution. This research deals with the estimation of scale parameter for Dirichlet distribution with known shapes. We examined three methods to estimate the parameters of Dirichlet distribution which are Maximum Likelihood Estimator, Method of Moment Estimator and Quasi-Likelihood Estimator. The performance of these methods were compared at different sample sizes using Bias, Mean Square Error, Mean Absolute Error and Variance criteria, an extensive simulation study was carried out on the basis of the selected criterion using statistical software packages as well as the application of the criterion to real life data, all these were done to obtain the most efficient method. The simulation study and analysis revealed that Quasi-Likelihood Estimator perform better in terms of bias while Method of Moment Estimator is better than the other two methods in terms of variance; the Maximum Likelihood Estimation was the best estimation method in terms of the Mean square Error and Mean Absolute Error; while Quasi-Likelihood Estimation method was the best estimation method with real life data.
Downloads
How to Cite
References
Murray Aitkin, Donald Rubin (1985) Estimation and Hypothesis Testing in Finite Mixture Models. 47(1), 67-75.
H Akaike (1973) Information theory and an extension of the maximum likelihood principle. 267-281.
C Andrieu, A Doucet (2001) Recursive Monte Carlo algorithms for parameter estimation in general state space models. 14-17.
Charles Antoniak (1974) Mixtures of Dirichlet Processes with Applications to Bayesian Nonparametric Problems. 2(6), 1152-1174.
Jeffrey Banfield, Adrian Raftery (1993) Model-Based Gaussian and Non-Gaussian Clustering. 49(3), 803.
C Biernacki, G Celeux, G Govaert (1999) An improvement of the NEC criterion for assessing the number of clusters in a mixture model. 20, 267-272.
C Biernacki, G Celeux, G Govaert (2000) Assessing a mixture model for clustering with the integrated completed likelihood. 22(7), 719-725.
Kai Ng, Guo‐liang Tian, Man‐lai Tang (2011) Dirichlet and Related Distributions. 888.
Robert Kass, Adrian Raftery (1995) Bayes Factors. 90(430), 773-795.
Ken Aho, Dewayne Derryberry, Teri Peterson (2014) Model selection for ecologists: the worldviews of AIC and BIC. 95(3), 631-636.
Z Liu, J Almhana, V Choulakian, R Mcgorman (2005) Online EM algorithm for mixture with application to internet traffic modeling. 50(4), 1052-1071.
A Mario, Anil Figueiredo, Jain (2002) Unsupervised learning of finite mixture models. 24(3), 381-396.
R Maitra (2009) Initializing Partition-Optimization Algorithms. 6(1), 144-157.
Benjamin Marlin, Richard Zemel (2004) The multiple multiplicative factor model for collaborative filtering. 73.
R Maitra, V Melnykov (2010) Assessing significance in finite mixture models.
R Maitra, V Melnykov (2010) Simulating data to study performance of finite mixture modeling and clustering algorithms.
G Mclachlan, T Krishnan (1997) The EM Algorithm and Extensions.
Michael Escobar, Mike West (1995) Bayesian Density Estimation and Inference Using Mixtures. 90(430), 577-588.
Narayanaswamy Balakrishnan, Valery Nevzorov (2004) A Primer on Statistical Distributions.
R Neal (1991) Bayesian mixture modeling. 197-211.
Nicholas T Longford (1987) A fast scoring algorithm for maximum likelihood estimation in unbalanced mixed models with nested random effects. 74(4), 817-827.
Nizar Bouguila, Djemel Ziou (2006) A hybrid SEM algorithm for high-dimensional unsupervised learning using a finite generalized Dirichlet mixture. 15(9), 2657-2668.
Nizar Bouguila, Djemel Ziou, Jean Vaillancourt (2004) Unsupervised Learning of a Finite Mixture Model Based on the Dirichlet Distribution and Its Application. 13(11), 1533-1543.
K Pearson (1894) Contribution to the mathematical theory of evolution. 185, 71-110.
Kathryn Roeder, Larry Wasserman (1997) Practical Bayesian Density Estimation Using Mixtures of Normals. 92(439), 894-902.
Surajit Ray, Bruce Lindsay (2008) Model Selection in High Dimensions: A Quadratic-Risk-Based Approach. 70(1), 95-118.
Masa-Aki Sato, Shin Ishii (2000) On-line EM Algorithm for the Normalized Gaussian Network. 12(2), 407-432.
S Choi, R Wette (1969) Maximum Likelihood Estimation of the Parameters of the Gamma Distribution and Their Bias. 11(4), 683-690.
Gideon Schwarz (1978) Estimating the Dimension of a Model. 6(2), 461-464.
S Maceachern (1999) Dependent nonparametric processes.
Jayaram Sethuraman, Ram Tiwari (1982) CONVERGENCE OF DIRICHLET MEASURES AND THE INTERPRETATION OF THEIR PARAMETER11Research supported by the U.S. Army Research Office under Grant no. DAAG 29-79-C-0158. The United States Government is authorized to reproduce and distribute reprints for governmental purposes.. 2, 305-315.
K Yeung, C Fraley, A Murua, A Raftery, W Ruzzo (2001) Model-based clustering and data transformations for gene expression data. 17(10), 977-987.
Z Zivkovic, F Van Der Heijden (2004) Recursive unsupervised learning of finite mixture models. 26(5), 651-656.
Zoran Zivkovic, Ferdinand Van Der Heijden (2004) Recursive unsupervised learning of finite mixture models. 26(5), 651-656.
Zoubin Ghahramani (2004) Unsupervised learning. 72-112.
Published
2020-09-30
Issue
Section
License
Copyright (c) 2020 Authors and Global Journals Private Limited
This work is licensed under a Creative Commons Attribution 4.0 International License.