Finite mixture models have come to play a very prominent role in modelling data. The finite mixture model is predicated on the assumption that distinct latent groups exist in the population. The finite mixture model therefore is based on a categorical latent variable that distinguishes the different groups. Often in practice, distinct sub-populations do not actually exist. For example, disease severity (e.g., depression) may vary continuously and therefore, a distinction of diseased and non-diseased may not be based on the existence of distinct sub-populations. Thus, what is needed is a generalization of the finite mixture’s discrete latent predictor to a continuous latent predictor. We cast the finite mixture model as a regression model with a latent Bernoulli predictor. A latent regression model is proposed by replacing the discrete Bernoulli predictor by a continuous latent predictor with a beta distribution. Motivation for the latent regression model arises from applications where distinct latent classes do not exist, but instead individuals vary according to a continuous latent variable. The shapes of the beta density are very flexible and can approximate the discrete Bernoulli distribution. Examples and a simulation are provided to illustrate the latent regression model. In particular, the latent regression model is used to model placebo effect among drug-treated subjects in a depression study.
Aitchison J and Shen S (1980) Logistic normal distributions: some properties and uses . Biometrika, 67, 261–72 .
2.
Azzalini A and Capitanio A (1999) Statistical applications of the multivariate skew-normal distribution . Journal of the Royal Statistical Society, Series B, 61, 579–602 .
3.
Bandeen-Roche K , Miglioretti D , Zeger S and Rathouz P (1997) Latent variable regression for multiple discrete outcomes . Journal of the American Statistical Association, 92, 1375–86 .
4.
Bartoletti S , Flury B and Nel D (1999). Allometric extension . Biometrika, 55, 1210–14 .
5.
Bartolucci F and Scaccia L (2005) The use of mixtures for dealing with non-normal regression errors . Computational Statistics & DataAnalysis, 48, 821–48 .
6.
Blei DM , Ng AY and Jordan M (2003) Latent dirichlet allocation . Journal of Machine Learning, 3, 993–1022 .
7.
Brown WA , Johnson MF and Chen MG (1992) Clinical features of depressed patients who do and do not improve with placebo . Psychiatry Research, 41, 203–14 .
8.
Bruni C and Koch G (1985) Identifiability of continuous mixtures of unknown gaussian distributions . Annals of Probability, 13, 1341–57 .
9.
Byrd RH , Lu P , Nocedal J and Zhu C (1995) A limited memory algorithm for bound constrained optimization . SIAM Journal of Scientific Computing, 16, 1190–208 .
10.
Day NE (1969) Estimating the components of a mixture of normal distributions . Biometrika, 56, 463–74 .
11.
Dempster AP , Laird NM and Rubin DB (1977) Maximum likelihood from incomplete data via the EM algorithm . Journal of the American Statistical Association, 39, 1–38 .
12.
Dorazio RM and Royle JA (2003) Mixture models for estimating the size of a closed population when capture rates vary among individuals . Biometrics, 59, 351–64 .
13.
Erosheva EA (2003) Bayesian estimation of the grade of membership model, pp. 501–10. Oxford: Oxford University Press .
14.
Everitt BS (1984) Maximum likelihood estimation of the parameters in a mixture of two univariate normal distributions; a comparison of different algorithms . The Statistician, 33, 205–15 .
15.
Fraley C and Raftery AE (2002) Model-based clustering, discriminant analysis and density estimation . Journal of the American Statistical Association, 9, 611–31 .
16.
Goodman LA (1974) Exploratory latent structure analysis using both identifiable and unidentifiable models . Biometrika, 61, 215–31 .
17.
Guo J , Wall M and Amemiya Y (2006) Latent class regression on latent factors . Biostatistics, 7, 145–63 .
18.
Gupta A and Wong C (1985) On three and five parameter bivariate beta distributions . Metrika, 32, 85–91 .
19.
Hartigan JA and Wong MA (1979) A k-means clustering algorithm . Applied Statistics, 28, 100–108 .
20.
Hastie T , Tibshirani R and Friedman J (2001) The elements of statistical learning. New York: Springer .
21.
James G and Sugar C (2003) Clustering for sparsely sampled functional data . Journal of the American Statistical Association, 98, 397–408 .
22.
Johnson NL , Kotz S and Read CB (1988) Skew-normal distributions. In Kotz S , Read CB and Banks DL (eds). Encyclopedia of statistical sciences, New York: Wiley , pp. 507–07.
23.
Johnson RA and Wichern DW (2005) Applied multivariate statistical analysis. New York: Pearson .
24.
Khan A and Brown WA (2001) The placebo enigma in antidepressant clinical trials . Journal of Clinical Psychopharmacology, 21, 123–25 .
25.
Leisch F (2004) Flexmix: a general framework for finite mixture models and latent class regression in r . Journal of Statistical Software, 11, 1–18 .
26.
MacQueen J (1967) Some methods for classification and analysis of multivariate observations . Proceedings 5th Berkeley Symposium on Mathematics, Statistics and Probability, 3, 281–97 .
27.
McGrath PJ , Stewart JW , Petkova E , Quitkin FM , Amsterdam JD , Fawcett J , Reimherr FW , Rosenbaum JF and Beasley CM (2000) Predictors of relapse during fluoxetine continuation or maintenance treatment for major depression . Journal of Clinical Psychiatry, 61, 518–24 .
28.
McLachlan GJ and Krishnan T (1997) The EM algorithm and extensions. New York: Wiley .
29.
Moustaki I and Knott M (2000) Generalized latent trait models . Psychometrika, 65, 391–411 .
30.
Mulrow CD , Williams JW , Trivedi M , Chiquette E , Aguilar C , Cornell JE , Badgett R , Noel PH , Lawrence V , Lee S , Luther M , Ramirez G , Richardson WS , and Stamm K (1999) Treatment of depression: newer pharmacotherapies. Rockville, MD: Agency for Health care Policy and Research .
31.
Muthén B (2004) Latent variable analysis: growth mixture modeling and related techniques for longitudinal data. In Kaplan D (ed.) Handbook of quantitative methodology for the social sciences, pp. 345–68. Newbury Park, CA: Sage Publications .
32.
Muthén B and Shedden K (1999) Finite mixture modeling with mixture outcomes using the EM algorithm . Biometrics, 55, 463–69 .
33.
Olkin I and Liu R (2003) A bivariate beta distribution . Statistics & Probability Letters, 62, 407–12 .
34.
Potthoff RF , Manton KG and Woodbury MA (2000) Dirichlet generalizations of latent-class models . Journal of Classification, 17, 315–25 .
35.
R Development Core Team (2010) R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing .
36.
Rousseauw J , du Plessis J , Benade A , Jordann P , Kotze J , Jooste P and Fereira J (1983) Coronary risk factor screening in three rural communties . South African Medical Journal, 64, 430–36 .
37.
Shapiro SS and Wilks MB (1965) An analysis of variance test for normality (complete samples) . Biometrika, 52, 591–611 .
38.
Tarpey T and Ivey CT (2006) Allometric extension for multivariate regression models . Journal of Data Science, 4, 479–95 .
39.
Thase M (1999) How should efficacy be evaluated in randomized clinical trials of treatments for depression? . Journal of Clinical Psychiatry, 60, 23–31 .
40.
Thisted RA (1991) Elements of statistical computing. New York: Chapman & Hall .
41.
Titterington DM , Smith AFM and Makov UE (1985) Statistical analysis of finite mixture distributions. New York: Wiley .
42.
Walsh BT , Seidman SN , Sysko R and Gould M (2002) Placebo response in studies of major depression: variable, substantial, and growing . Journal of the American Medical Association, 287, 1840–47 .
43.
Wedel M and DeSarbo WS (1995) A mixture likelihood approach for generalized linear models . Journal of Classification, 12, 21–55 .
44.
Woodbury MA , Clive J , and Garson A (1978) Mathematical typology—grade of membership technique for obtaining disease definition . Computational Biomedical Research, 11, 277–98 .
