Breakdown points for maximum likelihood estimators of location-scale mixtures
From MaRDI portal
Abstract: ML-estimation based on mixtures of Normal distributions is a widely used tool for cluster analysis. However, a single outlier can make the parameter estimation of at least one of the mixture components break down. Among others, the estimation of mixtures of t-distributions by McLachlan and Peel [Finite Mixture Models (2000) Wiley, New York] and the addition of a further mixture component accounting for ``noise by Fraley and Raftery [The Computer J. 41 (1998) 578-588] were suggested as more robust alternatives. In this paper, the definition of an adequate robustness measure for cluster analysis is discussed and bounds for the breakdown points of the mentioned methods are given. It turns out that the two alternatives, while adding stability in the presence of outliers of moderate size, do not possess a substantially better breakdown behavior than estimation based on Normal mixtures. If the number of clusters s is treated as fixed, r additional points suffice for all three methods to let the parameters of r clusters explode. Only in the case of r=s is this not possible for t-mixtures. The ability to estimate the number of mixture components, for example, by use of the Bayesian information criterion of Schwarz [Ann. Statist. 6 (1978) 461-464], and to isolate gross outliers as clusters of one point, is crucial for an improved breakdown behavior of all three techniques. Furthermore, a mixture of Normals with an improper uniform distribution is proposed to achieve more robustness in the case of a fixed number of components.
Recommendations
Cites work
- A constrained formulation of maximum-likelihood estimation for normal mixture distributions
- A General Qualitative Definition of Robustness
- A maximum likelihood methodology for clusterwise linear regression
- A new look at the statistical model identification
- A review of semiparametric mixture models
- An entropy criterion for assessing the number of clusters in a mixture model
- Breakdown and groups. (With discussions and rejoinder)
- Breakdown properties of location \(M\)-estimators
- Consistent estimation of the order of mixture models.
- Estimating the components of a mixture of normal distributions
- Estimating the dimension of a model
- Finite mixture models
- Finite sample breakdown points of projection based multivariate location and scatter statistics
- How Many Clusters? Which Clustering Method? Answers Via Model-Based Cluster Analysis
- scientific article; zbMATH DE number 3810727 (Why is no real title available?)
- scientific article; zbMATH DE number 3829050 (Why is no real title available?)
- scientific article; zbMATH DE number 3942788 (Why is no real title available?)
- scientific article; zbMATH DE number 3567782 (Why is no real title available?)
- scientific article; zbMATH DE number 194758 (Why is no real title available?)
- scientific article; zbMATH DE number 845707 (Why is no real title available?)
- Mixture Densities, Maximum Likelihood and the EM Algorithm
- Mixture Models, Robustness, and the Weighted Likelihood Methodology
- Model-Based Gaussian and Non-Gaussian Clustering
- Nearest-Neighbor Clutter Removal for Estimating Features in Spatial Point Processes
- Practical Bayesian Density Estimation Using Mixtures of Normals
- Robust Estimation of a Location Parameter
- Robust Statistics
- Robustness in statistical pattern recognition. Transl. from the Russian. Updated and exp. transl
- Robustness of ML Estimators of Location-Scale Mixtures
- Robustness Properties of k Means and Trimmed k Means
- Some quantitative relationships between two types of finite sample breakdown point
- The Identification of Multiple Outliers
- The Influence Curve and Its Role in Robust Estimation
Cited in
(60)- Extension of the mixture of factor analyzers model to incorporate the multivariate \(t\)-distribution
- Cluster-wise assessment of cluster stability
- Robust fitting of mixtures using the trimmed likelihood estimator
- Robust fitting of mixture regression models
- Robust mixture regression using the \(t\)-distribution
- Inference and mixture modeling with the elliptical Gamma distribution
- The joint role of trimming and constraints in robust estimation for mixtures of Gaussian factor analyzers
- Unsupervised learning of mixture regression models for longitudinal data
- A reweighting approach to robust clustering
- Robust, fuzzy, and parsimonious clustering, based on mixtures of factor analyzers
- Breakdown properties of location \(M\)-estimators
- On a resampling approach for tests on the number of clusters with mixture model-based clustering of tissue samples
- Computational aspects of fitting mixture models via the expectation-maximization algorithm
- Clustering for multivariate continuous and discrete longitudinal data
- Semi-supervised nonparametric Bayesian modelling of spatial proteomics
- Estimation and computations for Gaussian mixtures with uniform noise under separation constraints
- Mixtures of multivariate contaminated normal regression models
- Outlier identification in model-based cluster analysis
- Robust clustering in regression analysis via the contaminated Gaussian cluster-weighted model
- Eigenvalues and constraints in mixture modeling: geometric and computational issues
- Robust model-based clustering via mixtures of skew-t distributions with missing information
- A review of robust clustering methods
- A simulation study to compare robust clustering methods based on mixtures
- Robust estimation in the normal mixture model
- Model-based clustering
- Mixtures of autoregressions with an improper component for panel data
- Parsimonious mixtures of multivariate contaminated normal distributions
- Robust fitting of mixtures of factor analyzers using the trimmed likelihood estimator
- Detecting wine adulterations employing robust mixture of factor analyzers
- Issues of robustness and high dimensionality in cluster analysis
- A fast algorithm for robust constrained clustering
- Robust Estimation in the Normal Mixture Model Based on Robust Clustering
- On the breakdown behavior of the TCLUST clustering procedure
- Mixtures of regression models with incomplete and noisy data
- Wild adaptive trimming for robust estimation and cluster analysis
- Consistency, breakdown robustness, and algorithms for robust improper maximum likelihood clustering
- Latent simplex position model: high dimensional multi-view clustering with uncertainty quantification
- Fitting insurance and economic data with outliers: a flexible approach based on finite mixtures of contaminated gamma distributions
- Italian contributions on some recent research topics in cluster analysis
- Mixture of linear mixed models using multivariate t distribution
- Robustness of ML Estimators of Location-Scale Mixtures
- Maximum likelihood estimation in constrained parameter spaces for mixtures of factor analyzers
- Avoiding spurious local maximizers in mixture modeling
- An adequacy approach for deciding the number of clusters for OTRIMLE robust Gaussian mixture‐based clustering
- Graphical and Computational Tools to Guide Parameter Choice for the Cluster Weighted Robust Model
- Exploring the number of groups in robust model-based clustering
- Robust clustering based on trimming
- Parsimonious seemingly unrelated contaminated normal cluster-weighted models
- t distribution-based robust semiparametric mixture regression model
- A refreshing take on the inverted Dirichlet via a mode parameterization with some statistical illustrations
- Finite mixture of regression models for censored data based on the skew-t distribution
- High breakdown mixture discriminant analysis
- Maximum likelihood estimation of heterogeneous mixtures of Gaussian and uniform distributions
- A GARCH-type model with cross-sectional volatility clusters
- Estimating finite mixture models using component self-paced learning
- S-estimation of hidden Markov models
- On the breakdown point of multivariate location estimators based on trimming procedures
- Dissolution point and isolation robustness: Robustness criteria for general cluster analysis methods
- A general trimming approach to robust cluster analysis
- Breakdown concepts for contingency tables
This page was built for publication: Breakdown points for maximum likelihood estimators of location-scale mixtures
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1879957)