Markov Chain Monte Carlo Methods and the Label Switching Problem in Bayesian Mixture Modeling

By Ajay Jasra, Christopher C. Holmes, David A. Stephens. 2005

721 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/d040fbb14ac4cdd42dbb154e30eb3fcba3967dbc

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistics and Probability

Citations in the corpus

An Introduction to Bayesian Inference via Variational Approximations (2011, methodology): “Current recommendations are to run a chain without constraints and then postprocess using a variety of methods (Jasra, Holmes, and Stephens 2005).”
CLUMPP: a cluster matching and permutation program for dealing with label switching and multimodality in analysis of population structure (2007, background): “‘Label switching’ refers to a scenario in which different replicates obtain the same membership coefficient estimates, except with a different permutation of the cluster labels (Jasra et al. , 2005; Stephens, 2000).”
Model-based clustering of categorical time series (2010, methodology): “Like for any finite mixture model, label switching may occur during MCMC sampling both for Markov chain clustering as well as for Dirichlet multinomial clustering, see Jasra et al. (2005) or Frühwirth-Schnatter (2006, Section 3.5) for a recent review.”
A Sticky HDP-HMM With Application to Speaker Diarization (2009, background): “Jasra et al. (2005) provides an overview of some related techniques to address the label-switching issue.”
Penalized Model-Based Clustering with Application to Variable Selection (2007, background): “Recently Fraley and Raftery (2005) proposed a Bayesian approach to regularizing model-based clustering; there is a large body of literature on Bayesian mixture modeling, e.g., Richardson and Green (1997), Jasra et al. (2005) and references therein.”
Skew mixture models for loss distributions: a Bayesian approach (2012, background): “(2.6) The Gaussian prior for ξl is taken to be rather flat over the corresponding observed range of the data (R) as in Jasra et al. (2005); in the application considered in the next section we take ζ = R/2, and κ = R.”
Model-Based Clustering of Non-Gaussian Panel Data Based on Skew-t Distributions (2010, background): “Comprehensive discussions have been given by Jasra, Holmes, and Stephens (2005) and Frühwirth-Schnatter (2006).”
Testing hypotheses via a mixture estimation model (2014, background): “Another computational challenge that plagues Bayesian estimation for most mixture models is “label switching” (Celeux et al., 2000; Stephens, 2000; Jasra et al., 2005).”
GFlowNet Foundations (2021, neutral): “This is known as the mode-mixing problem (Jasra et al., 2005; Bengio et al., 2013; Pompe et al., 2020): the chances of going from one mode to a neighboring one may become exponentially small (and thus require exponentially long chains) if the modes are separated by a long sequence of…”
Rapid changes in growth, condition, size and age of small pelagic fish in the Mediterranean (2014, positive): “For our data, this improved the results significantly as label switching (when within or between 175 chains labels switch between peaks) was avoided, allowing the model to converge (Stephens 2000; Jasra et al. 2005).”
Estimating the Marginal Likelihood Using the Arithmetic Mean Identity (2017, background): “Meng and Wong (1996), Raftery (1996), Lewis and Raftery (1997), Han and Carlin (2001), Raftery et al.”