×

zbMATH — the first resource for mathematics

Bayesian factor analysis with fat-tailed factors and its exact marginal likelihood. (English) Zbl 1163.62044
Summary: The traditional Bayesian factor analysis method is extended. In contrast to the case for previous studies, the matrix variate \(t\)-distribution is utilized to provide a prior density on the latent factors. This is a natural extension of the traditional model and yields many advantages. The crucial issue is the selection of the number of factors. The marginal likelihood, constructed by asymptotic and computational approaches, is generally utilized for this problem. However, both theoretical and computational problems have arisen.
The exact marginal likelihood is derived. It enables us to evaluate posterior model probabilities without inducing the above problems. Monte Carlo experiments were conducted to examine the performance of the proposed Bayesian factor analysis modelling methodology. The simulation results show that the proposed methodology performs well.

MSC:
62H25 Factor analysis and principal components; correspondence analysis
62F15 Bayesian inference
65C60 Computational problems in statistics (MSC2010)
65C05 Monte Carlo methods
Software:
tsbridge
PDF BibTeX XML Cite
Full Text: DOI
References:
[1] Aguilar, O.; West, M., Bayesian dynamic factor models and variance matrix discounting for portfolio allocation, J. bus. econom. statist., 18, 338-357, (2000)
[2] Lopes, H.F.; West, M., Bayesian model assessment in factor analysis, Statist. sinica, 14, 41-67, (2004) · Zbl 1035.62060
[3] Press, S.J.; Shigemasu, K., Bayesian inference in factor analysis, (), 271-287
[4] West, M., Bayesian factor regression models in the “large p, small n” paradigm, (), 723-732
[5] Kass, R.; Raftery, A., Bayes factors, J. amer. statist. assoc., 90, 773-795, (1995) · Zbl 0846.62028
[6] Press, S.J.; Shigemasu, K., A note on choosing the number of factors, Comm. statist. theory methods, 28, 1653-1670, (1999) · Zbl 0937.62060
[7] Schwarz, G., Estimating the dimension of a model, Ann. statist., 6, 461-464, (1978) · Zbl 0379.62005
[8] Konishi, S.; Ando, T.; Imoto, S., Bayesian information criteria and smoothing parameter selection in radial basis function networks, Biometrika, 91, 27-43, (2004) · Zbl 1132.62313
[9] Berger, J.O.; Ghosh, J.K.; Mukhopadhyay, N., Approximation and consistency of Bayes factors as model dimension grows, J. statist. plann. inference, 112, 241-258, (2003) · Zbl 1026.62018
[10] Lee, S.Y.; Song, X.Y., Bayesian selection on the number of factors in a factor analysis model, Behaviormetrika, 29, 23-39, (2002) · Zbl 1014.62034
[11] Press, S.J., Subjective and objective Bayesian statistics. principles models and applications, (2003), Wiley New Jersey · Zbl 1014.62028
[12] Chib, S., Marginal likelihood from the Gibbs output, J. amer. statist. assoc., 90, 1313-1321, (1995) · Zbl 0868.62027
[13] Newton, M.A.; Raftery, A.E., Approximate Bayesian inference with the weighted likelihood bootstrap, J. roy. stat. soc. ser. B, 56, 3-48, (1994) · Zbl 0788.62026
[14] Gelfand, A.E.; Dey, D., Bayesian model choice: asymptotics and exact calculations, J. roy. stat. soc. ser. B, 56, 501-514, (1994) · Zbl 0800.62170
[15] Lewis, S.M.; Raftery, A.E., Estimating Bayes factors via posterior simulation with the laplace – metropolis estimator, J. amer. statist. assoc., 92, 648-655, (1997) · Zbl 0889.62018
[16] Meng, X.L.; Wong, W.H., Simulating ratios of normalizing constants via a simple identity: A theoretical exploration, Statist. sinica, 6, 831-860, (1996) · Zbl 0857.62017
[17] Green, P., Reversible jump Markov chain Monte Carlo computation and Bayesian model determination, Biometrika, 82, 711-732, (1995) · Zbl 0861.62023
[18] Tucker, L.R., The role of correlated factors in factor analysis, Psychometrika, 5, 141-152, (1940) · JFM 66.0653.06
[19] Lee, S.Y.; Xia, Y.M., Maximum likelihood methods in treating outliers and symmetrically heavy-tailed distributions for nonlinear structural equation models with missing data, Psychometrika, 71, 565-585, (2006) · Zbl 1306.62462
[20] Lee, S.Y.; Xia, Y.M., A robust Bayesian approach for structural equation models with missing data, Psychometrika, 73, 343-364, (2008) · Zbl 1301.62122
[21] Lee, S.Y.; Xia, Y.M., Semiparametric Bayesian analysis of structural equation models with fixed covariates, Statist. med., 27, 2341-2360, (2008)
[22] Geweke, J.F., Contemporary Bayesian econometrics and statistics, (2005), Wiley New York · Zbl 1093.62107
[23] Lawley, D.N., The estimation of factor loadings by the method of maximum likelihood, Proc. roy. soc. Edinburgh ser. A, 60, 64-82, (1940) · Zbl 0027.23503
[24] Seber, G.A.F., Multivariate observations, (1984), Wiley New York · Zbl 0627.62052
[25] Geweke, J.F.; Singleton, K.J., Interpreting the likelihood ratio statistic in factor models when sample size is small, J. amer. statist. assoc., 75, 133-137, (1980) · Zbl 0424.62039
[26] Geweke, J.F.; Zhou, G., Measuring the pricing error of the arbitrage pricing theory, Rev. finan. stud., 9, 557-587, (1996)
[27] Yalcin, I.; Amemiya, Y., Nonlinear factor analysis as a statistical method, Statist. sci., 16, 275-294, (2001) · Zbl 1059.62552
[28] Lewin-Koh, S.C.; Amemiya, Y., Heteroscedastic factor analysis, Biometrika, 90, 85-97, (2003) · Zbl 1034.62051
[29] D.B. Rowe, Correlated Bayesian factor analysis, Ph.D. Thesis, Department of Statistics, University of California, Riverside, 1998
[30] Dickey, J.M., Matricvariate generalizations of the multivariate t-distribution and the inverted multivariate t-distribution, Ann. math. statist., 38, 511-518, (1967) · Zbl 0158.18403
[31] Javier, W.R.; Gupta, A.K., On matric variate-\(t\) distribution, Comm. statist. theory methods, 14, 1413-1425, (1985) · Zbl 0577.62045
[32] Phillips, P.C.B., The distribution of matrix quotients, J. multivariate anal., 16, 157-161, (1985) · Zbl 0558.62045
[33] Lee, S.Y., Structural equation modelling: A Bayesian approach, (2007), John Wiley and Sons England
[34] Ando, T., Bayesian inference for nonlinear and non-Gaussian stochastic volatility model with leverage effect, J. Japan statist. soc., 36, 173-197, (2006) · Zbl 1110.62141
[35] Ando, T., Bayesian predictive information criterion for the evaluation of hierarchical Bayesian and empirical Bayes models, Biometrika, 94, 443-458, (2007) · Zbl 1132.62005
[36] Ando, T., Bayesian model averaging and Bayesian predictive information criterion for model selection, J. Japan statist. soc., 38, 243-257, (2008)
[37] Ando, T.; Konishi, S.; Imoto, S., Nonlinear regression modeling via regularized radial basis function networks, J. statist. plann. inference, 138, 3616-3633, (2008) · Zbl 1152.62039
[38] T. Ando, S. Konishi, Nonlinear logistic discrimination via regularized radial basis functions for classifying high-dimensional data, Ann. Inst. Statist. Math. (2007) (in press) · Zbl 1332.62198
[39] Zellner, A., An introduction to Bayesian inference in econometrics, (1971), John Wiley and Sons New York · Zbl 0246.62098
[40] Konishi, S.; Kitagawa, G., Generalised information criteria in model selection, Biometrika, 83, 875-890, (1996) · Zbl 0883.62004
[41] Gelfand, A.E.; Dey, D.; Chang, H., Model determination using predictive distributions with implementation via sampling-based methods (with discussion), (), 147-167
[42] Rossi, P.; Gilula, Z.; Allenby, G., Overcoming scale usage heterogeneity: A Bayesian hierarchical approach, J. amer. statist. assoc., 96, 20-31, (2001)
[43] Kendall, M., Multivariate analysis, (1980), Charles Griffin and Company LTD. London
[44] Bai, J.; Ng, S., Determining the number of factors in approximate factor models, Econometrica, 70, 191-221, (2002) · Zbl 1103.91399
[45] Akaike, H., Factor analysis, AIC psychometrika, 3, 317-332, (1987) · Zbl 0627.62067
[46] Bozdogan, H., Model selection and akaike’s information criterion (AIC): the general theory and its analytical extensions, Psychometrika, 52, 345-370, (1987) · Zbl 0627.62005
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.