Gaussian parsimonious clustering models with covariates and a noise component

DOI10.1007/S11634-019-00373-8zbMATH Open1474.62240arXiv1711.05632OpenAlexW2973763519WikidataQ127226156 ScholiaQ127226156MaRDI QIDQ84391FDOQ84391

Authors: Keefe Murphy, Keefe Murphy, Thomas Brendan Murphy 

Publication date: 20 September 2019

Published in: Advances in Data Analysis and Classification. ADAC (Search for Journal in Brave)

Abstract: We consider model-based clustering methods for continuous, correlated data that account for external information available in the presence of mixed-type fixed covariates by proposing the MoEClust suite of models. These models allow different subsets of covariates to influence the component weights and/or component densities by modelling the parameters of the mixture as functions of the covariates. A familiar range of constrained eigen-decomposition parameterisations of the component covariance matrices are also accommodated. This paper thus addresses the equivalent aims of including covariates in Gaussian parsimonious clustering models and incorporating parsimonious covariance structures into all special cases of the Gaussian mixture of experts framework. The MoEClust models demonstrate significant improvement from both perspectives in applications to both univariate and multivariate data sets. Novel extensions to include a uniform noise component for capturing outliers and to address initialisation of the EM algorithm, model selection, and the visualisation of results are also proposed.

Full work available at URL: https://arxiv.org/abs/1711.05632

Recommendations

• Stable and visualizable Gaussian parsimonious clustering models
• Model-based clustering with sparse covariance matrices
• Using conditional independence for parsimonious model-based Gaussian clustering
• Model-based clustering with noise: Bayesian inference and estimation
• Mixtures of general location model with factor analyzer covariance structure for clustering mixed type data

zbMATH Keywords

EM algorithmmodel-based clusteringcovariatesmixtures of expertsmultivariate responsenoise componentparsimony

Mathematics Subject Classification ID

Factor analysis and principal components; correspondence analysis (62H25) Classification and discrimination; cluster analysis (statistical aspects) (62H30) Generalized linear models (logistic models) (62J12)

Cites Work

• Fitting finite mixtures of generalized linear regressions in \textsf{R}
• Model based clustering for mixed data: clustMD
• Estimating the dimension of a model
• Title not available (Why is that?)
• The distribution of the likelihood ratio for mixtures of densities from the one-parameter exponential family
• Parsimonious mixtures of multivariate contaminated normal distributions
• Title not available (Why is that?)
• Model-Based Gaussian and Non-Gaussian Clustering
• Pattern recognition and machine learning.
• An Efficient Method of Estimating Seemingly Unrelated Regressions and Tests for Aggregation Bias
• Bayesian regularization for normal mixture estimation and model-based clustering
• Identifiability of models for clusterwise linear regression
• The generalized linear mixed cluster-weighted model
• Multivariate response and parsimony for Gaussian cluster-weighted models
• Eigenvalues and constraints in mixture modeling: geometric and computational issues
• Mixtures of regressions with predictor-dependent mixing proportions
• Local statistical modeling via a cluster-weighted approach with elliptical distributions
• Smoothly mixing regressions
• Clustering bivariate mixed-type data via the cluster-weighted model
• Finite Mixture Models with Concomitant Information: Assessing Diagnostic Criteria for Diabetes
• Mixed Poisson Regression Models with Covariate Dependent Rates
• Title not available (Why is that?)
• Title not available (Why is that?)
• Model-based clustering, classification, and discriminant analysis via mixtures of multivariate \(t\)-distributions

Cited In (12)

• Stable and visualizable Gaussian parsimonious clustering models
• Merging components in linear Gaussian cluster-weighted models
• Semiparametric finite mixture of regression models with Bayesian P-splines
• Composite likelihood methods for parsimonious model-based clustering of mixed-type data
• Conditional functional clustering for longitudinal data with heterogeneous nonlinear patterns
• Modeling the duration and size of wildfires using joint mixture models
• Addressing overfitting and underfitting in Gaussian model-based clustering
• Model-based clustering with noise: Bayesian inference and estimation
• MoEClust
• Multivariate cluster weighted models using skewed distributions
• A similarity-based Bayesian mixture-of-experts model
• Parsimonious mixture-of-experts based on mean mixture of multivariate normal distributions

Uses Software

• mclust
• R
• flexmix
• mixtools
• Rmixmod
• flexCWM
• clustMD
• MoEClust
• PRMLT

This page was built for publication: Gaussian parsimonious clustering models with covariates and a noise component

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q84391)