On the identifiability of Bayesian factor analytic models
From MaRDI portal
Publication:113012
DOI10.1007/S11222-022-10084-4zbMATH Open1484.62008arXiv2004.05105OpenAlexW3015764491MaRDI QIDQ113012FDOQ113012
Authors: Panagiotis Papastamoulis, Ioannis Ntzoufras, I. Ntzoufras 
Publication date: 27 February 2022
Published in: Statistics and Computing (Search for Journal in Brave)
Abstract: A well known identifiability issue in factor analytic models is the invariance with respect to orthogonal transformations. This problem burdens the inference under a Bayesian setup, where Markov chain Monte Carlo (MCMC) methods are used to generate samples from the posterior distribution. We introduce a post-processing scheme in order to deal with rotation, sign and permutation invariance of the MCMC sample. The exact version of the contributed algorithm requires to solve assignment problems per (retained) MCMC iteration, where denotes the number of factors of the fitted model. For large numbers of factors two approximate schemes based on simulated annealing are also discussed. We demonstrate that the proposed method leads to interpretable posterior distributions using synthetic and publicly available data from typical factor analytic models as well as mixtures of factor analyzers. An R package is available online at CRAN web-page.
Full work available at URL: https://arxiv.org/abs/2004.05105
Recommendations
- Dealing with reflection invariance in Bayesian factor analysis
- A Mode-Jumping Algorithm for Bayesian Factor Analysis
- Bayesian analysis of static and dynamic factor models: an ex-post approach towards the rotation problem
- Order-invariant prior specification in Bayesian factor analysis
- Bayesian exploratory factor analysis
Computational methods for problems pertaining to statistics (62-08) Bayesian inference (62F15) Factor analysis and principal components; correspondence analysis (62H25)
Cites Work
- Bayesian exploratory factor analysis
- Clustering multivariate data using factor analytic Bayesian mixtures with an unknown number of components
- Overfitting Bayesian mixtures of factor analyzers with an unknown number of components
- Estimating the dimension of a model
- Inference from iterative simulation using multiple sequences
- From here to infinity: sparse finite versus Dirichlet process mixtures in model-based clustering
- Title not available (Why is that?)
- Optimization by simulated annealing
- Reversible jump Markov chain Monte Carlo computation and Bayesian model determination
- Sparse Bayesian infinite factor models
- Latent variable models and factor analysis. A unified approach
- Dealing With Label Switching in Mixture Models
- Mixture model-based classification
- Model-based clustering based on sparse finite Gaussian mixtures
- Modelling high-dimensional data by mixtures of factor analyzers
- Bayesian computation and stochastic systems. With comments and reply.
- Sampling-Based Approaches to Calculating Marginal Densities
- Arbitrage, Factor Structure, and Mean-Variance Analysis on Large Asset Markets
- Title not available (Why is that?)
- Title not available (Why is that?)
- A joint treatment of varimax rotation and the problem of diagonalizing symmetric matrices simultaneously in the least-squares sense
- Characteristic polynomials of a permutation representation
- Generic global identification in factor analysis
- Mixtures of factor analysers. Bayesian estimation and inference by stochastic simulation
- Infinite mixtures of infinite factor analysers
- A Bayesian approach to nonlinear latent variable models using the Gibbs sampler and the Metropolis-Hastings algorithm
- A generalized solution of the orthogonal Procrustes problem
- A matrix formulation of Kaiser's varimax criterion
- A theoretical framework for simulated annealing
- Title not available (Why is that?)
- Assignment Problems
- Probabilistic Principal Component Analysis
- On finite sequences of real numbers
- On Bayesian model and variable selection using MCMC
- Stochastic search item selection for factor analytic models
- Expandable factor analysis
- High-dimensional sparse factor modeling: applications in gene expression genomics
- An Algorithm for the Traveling Salesman Problem
- A flexible AFT model for misclassified clustered interval-censored data
- A Mode-Jumping Algorithm for Bayesian Factor Analysis
- Bayesian analysis of static and dynamic factor models: an ex-post approach towards the rotation problem
- The varimax criterion for analytic rotation in factor analysis
Cited In (16)
- Title not available (Why is that?)
- Bayesian estimation of sparse dynamic factor models with order-independent and ex-post mode identification
- Bayesian analysis of the factor model with finance applications
- Post-processing of Markov chain Monte Carlo output in Bayesian latent variable models with application to multidimensional scaling
- Determining the number of factors in constrained factor models via Bayesian information criterion
- Dealing with reflection invariance in Bayesian factor analysis
- A sparse factor model for clustering high-dimensional longitudinal data
- Bayesian generalized linear low rank regression models for the detection of vaccine-adverse event associations
- factor.switching
- Graph link prediction in computer networks using Poisson matrix factorisation
- Invariant inference and efficient computation in the static factor model
- Title not available (Why is that?)
- Bayesian inference for graphical factor analysis models
- Bayesian factor selection in a hybrid approach to confirmatory factor analysis
- A note on identifiability conditions in confirmatory factor analysis
- A Bayesian semiparametric factor analysis model for subtype identification
Uses Software
This page was built for publication: On the identifiability of Bayesian factor analytic models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q113012)
