plfm-package: Probabilistic Latent Feature Analysis

In plfm: Probabilistic Latent Feature Analysis

Probabilistic Latent Feature Analysis

Description

Functions for estimating disjunctive, conjunctive or additive probabilistic latent feature models on (aggregated) binary three-way data

Details

Package:	plfm
Type:	Package
Version:	2.2.2
Date:	2018-10-27
License:	GPL(>=2)
LazyLoad:	yes

Probabilistic latent feature models can be used to model three-way three-mode binary observations (e.g. persons who indicate for each of a number of products and for each of a set of attributes whether a product has a certain attribute). A basic probabilistic feature model (referred to as plfm) uses aggregated three-way three-mode binary data as input, namely the two-way two-mode frequency table that is obtained by summing the binary three-way three-mode data across persons. The basic probabilistic feature model (Maris, De Boeck and Van Mechelen, 1996) is based on the assumption that observations are statistically independent and that model parameters are homogeneous across persons. The plfm function can be used to locate the posterior mode(s) of basic probabilistic feature models, and to compute information criteria for model selection, and measures of statistical and descriptive model fit. The stepplfm function can be used to fit a series of disjunctive, conjunctive or additive basic probabilistic feature models with different number of latent features. In addition, the bayesplfm function can be used to compute a sample of the posterior distribution of the basic probabilistic feature model in the neigbourhood of a specific posterior mode.

Latent class extensions of the probabilistic feature model (referred to as LCplfm) take binary three-way three-mode observations as input. In contrast to the basic probabilistic feature model, latent class probabilistic feature models allow to model dependencies between (subsets of) observations (Meulders, De Boeck and Van Mechelen, 2003) and/or to account for heterogeneity in model parameters across persons (Meulders, Tuerlinckx, and Vanpaemel, 2013). The LCplfm function can be used to compute posterior mode estimates (of different types of) latent class probabilistic feature models as well as to compute information criteria for model selection, and measures of descriptive model fit. The stepLCplfm function can be used to compute a series of latent class probabilistic feature models with different numbers of latent features and latent classes.

To see the preferable citation of the package, type citation("plfm").

Author(s)

Michel Meulders

Maintainer: <[e-mail]>

References

Candel, M. J. J. M., and Maris, E. (1997). Perceptual analysis of two-way two-mode frequency data: probability matrix decomposition and two alternatives. International Journal of Research in Marketing, 14, 321-339.

Gelman, A., Van Mechelen, I., Verbeke, G., Heitjan, D. F., and Meulders, M. (2005). Multiple imputation for model checking: Completed-data plots with missing and latent data. Biometrics, 61, 74-85.

Maris, E., De Boeck, P., and Van Mechelen, I. (1996). Probability matrix decomposition models. Psychometrika, 61, 7-29.

Meulders, M. (2013). An R Package for Probabilistic Latent Feature Analysis of Two-Way Two-Mode Frequencies. Journal of Statistical Software, 54(14), 1-29. URL http://www.jstatsoft.org/v54/i14/.

Meulders, M., De Boeck, P., Kuppens, P., and Van Mechelen, I. (2002). Constrained latent class analysis of three-way three-mode data. Journal of Classification, 19, 277-302.

Meulders, M., De Boeck, P., and Van Mechelen, I. (2001). Probability matrix decomposition models and main-effects generalized linear models for the analysis of replicated binary associations. Computational Statistics and Data Analysis, 38, 217-233.

Meulders, M., De Boeck, P., and Van Mechelen, I. (2003). A taxonomy of latent structure assumptions for probability matrix decomposition models. Psychometrika, 68, 61-77.

Meulders, M., De Boeck, P., Van Mechelen, I., and Gelman, A. (2005). Probabilistic feature analysis of facial perception of emotions. Applied Statistics, 54, 781-793.

Meulders, M., De Boeck, P., Van Mechelen, I., Gelman, A., and Maris, E. (2001). Bayesian inference with probability matrix decomposition models. Journal of Educational and Behavioral Statistics, 26, 153-179.

Meulders, M. and De Bruecker, P. (2018). Latent class probabilistic latent feature analysis of three-way three-mode binary data. Journal of Statistical Software, 87(1), 1-45.

Meulders, M., Gelman, A., Van Mechelen, I., and De Boeck P. (1998). Generalizing the probability matrix decomposition model: An example of Bayesian model checking and model expansion. In J. Hox, and E. De Leeuw (Eds.), Assumptions, robustness, and estimation methods in multivariate modeling (pp. 1-19). TT Publicaties: Amsterdam.

Meulders, M., Tuerlinckx, F., and Vanpaemel, W. (2013). Constrained multilevel latent class models for the analysis of three-way three-mode binary data. Journal of Classification, 30 (3), 306-337.

plfm documentation built on March 30, 2022, 5:08 p.m.