Search

This paper presents a review about the usage of eigenvalues restrictions for constrained parameter estimation in mixtures of elliptical distributions according to the likelihood approach. These restrictions serve a twofold purpose: to avoid convergence to degenerate solutions and to reduce the onset of non interesting (spurious) maximizers, related to complex likelihood surfaces. The paper shows how the constraints may play a key role in the theory of Euclidean data clustering. The aim here is to provide a reasoned review of the constraints and their applications, along the contributions of many authors, spanning the literature of the last thirty years., Spanish Ministerio de Economía y Competitividad (grant MTM2017-86061-C2-1-P), Junta de Castilla y León - Fondo Europeo de Desarrollo Regional (grant VA005P17 and VA002G18)

Producción Científica, A robust estimator for a wide family of mixtures of linear regression is presented. Robustness is based on the joint adoption of the Cluster Weighted Model and of an estimator based on trimming and restrictions. The selected model provides the conditional distribution of the response for each group, as in mixtures of regression, and further supplies local distributions for the explanatory variables. A novel version of the restrictions has been devised, under this model, for separately controlling the two sources of variability identified in it. This proposal avoids singularities in the log-likelihood, caused by approximate local collinearity in the explanatory variables or local exact fits in regressions, and reduces the occurrence of spurious local maximizers. In a natural way, due to the interaction between the model and the estimator, the procedure is able to resist the harmful influence of bad leverage points along the estimation of the mixture of regressions, which is still an open issue in the literature. The given methodology defines a well-posed statistical problem, whose estimator exists and is consistent to the corresponding solution of the population optimum, under widely general conditions. A feasible EM algorithm has also been provided to obtain the corresponding estimation. Many simulated examples and two real datasets have been chosen to show the ability of the procedure, on the one hand, to detect anomalous data, and, on the other hand, to identify the real cluster regressions without the influence of contamination. Keywords Cluster Weighted Modeling · Mixture of Regressions · Robustness

A novel family of twelve mixture models with random covariates, nested in the linear $t$ cluster-weighted model (CWM), is introduced for model-based clustering. The linear $t$ CWM was recently presented as a robust alternative to the better known linear Gaussian CWM. The proposed family of models provides a unified framework that also includes the linear Gaussian CWM as a special case. Maximum likelihood parameter estimation is carried out within the EM framework, and both the BIC and the ICL are used for model selection. A simple and effective hierarchical random initialization is also proposed for the EM algorithm. The novel model-based clustering technique is illustrated in some applications to real data. Finally, a simulation study for evaluating the performance of the BIC and the ICL is presented.

Cluster-weighted modeling (CWM) is a mixture approach to modeling the joint probability of data coming from a heterogeneous population. Under Gaussian assumptions, we investigate statistical properties of CWM from both theoretical and numerical point of view; in particular, we show that Gaussian CWM includes mixtures of distributions and mixtures of regressions as special cases. Further, we introduce CWM based on Student-t distributions, which provides a more robust fit for groups of observations with longer than normal tails or noise data. Theoretical results are illustrated using some empirical studies, considering both simulated and real data. Some generalizations of such models are also outlined.

Mixtures of multivariate t-distributions provide a robust parametric extension to the fitting of data with respect to normal mixtures. In presence of some noise component, potential outliers or data with longer-than-normal tails, one way to broaden the model can be provided by considering $t$ distributions. In this framework, the degrees of freedom can act as a robustness parameter, tuning the heaviness of the tails, and down weighting the effect of the outliers on the parameters estimation. The aim of this paper is to extend to mixtures of multivariate elliptical distributions some theoretical results about the likelihood maximization on constrained parameter spaces. Further, a constrained monotone algorithm implementing maximum likelihood mixture decomposition of multivariate t-distributions is proposed, to achieve improved convergence capabilities and robustness. Monte Carlo numerical simulations and a real data study demonstrate the better performance of the algorithm, comparing it to earlier proposals

Cluster-weighted models (CWMs) are a flexible family of mixture models for fitting the joint distribution of a random vector composed of a response variable and a set of covariates. CWMs act as a convex combination of the products of the marginal distribution of the covariates and the conditional distribution of the response given the covariates. In this paper, we introduce a broad family of CWMs in which the component conditional distributions are assumed to belong to the exponential family and the covariates are allowed to be of mixed-type. Under the assumption of Gaussian covariates, sufficient conditions for model identifiability are provided. Moreover, maximum likelihood parameter estimates are derived using the EM algorithm. Parameter recovery, classification assessment, and performance of some information criteria are investigated through a broad simulation design. An application to real data is finally presented, with the proposed model outperforming other well-established mixture-based approaches.

Mixtures of factor analyzers are becoming more and more popular in the area of model based clustering of high-dimensional data. According to the likelihood approach in data modeling, it is well known that the unconstrained log-likelihood function may present spurious maxima and singularities and this is due to specific patterns of the estimated covariance structure, when their determinant approaches 0. To reduce such drawbacks, in this paper we introduce a procedure for the parameter estimation of mixtures of factor analyzers, which maximizes the likelihood function in a constrained parameter space. We then analyze and measure its performance, compared to the usual non-constrained approach, via some simulations and applications to real data sets., 21 pages, 7 figures, 8 tables

A new approach to structural equation modelling based on so-called Extended Redundancy Analysis has been recently proposed in literature, enhanced with the added characteristic of generalizing Redundancy Analysis and Reduced-Rank Regression models for more than two blocks. However, in presence of direct effects linking exogenous and endogenous variables, the latent composite scores are estimated by ignoring the presence of the specified direct effects. In this paper, we extend Extended Redundancy Analysis, permitting us to specify and fit a variety of relationships among latent composites and endogenous variables. In particular, covariates are allowed to affect endogenous indicators indirectly through the latent composites and/or directly.

Singular value decomposition (SVD) of skew-symmetric matrices was proposed by Gower (Recent Developments in Statistics, Amsterdam: North Holland, 1977) to represent asymmetry of proximity data. Some authors considered the plane (bimension or hedron) determined by the first two singular vectors to detect orderings (seriation) for preference or dominance data. Following these approaches, in this paper some procedures of asymmetric multidimensional scaling useful for seriation are proposed focalizing on a model that is a particular case of rank-2 SVD model. An application to Thurstone’s paired comparison data on the relative seriousness of crime is also presented.

A multivariate hidden Markov model is proposed for clustering mixed linear and circular time-series data with missing values. The model integrates von Mises and normal densities to describe the distribution that the data take under different latent regimes, with parameters that depend on the evolution of an unobserved Markov chain. Estimation is facilitated by an EM algorithm that treats the states of the latent chain and missing values as different sources of incomplete information. The model is exploited to identify sea regimes from multivariate marine data.

Cluster-Weighted Modeling is a flexible statistical framework for modeling local relationships in heterogeneous populations on the basis of weighted combinations of local models. Besides the traditional approach based on Gaussian assumptions, here we consider Cluster Weighted Modeling based on Student-t distributions. In this paper we present an EM algorithm for parameter estimation in Cluster-Weighted models according to the maximum likelihood approach. © 2012 Springer-Verlag Berlin Heidelberg.

This paper introduces the concept of the conditional impurity in the framework of tree-based models in order to deal with the analysis of three-way data, where a response variable and a set of predictors are measured on a sample of objects in different occasions. The conditional impurity in the definition of splitting criterion is defined as a classical impurity measure weighted by a predictability index.

We introduce the control sample in survey sampling as a tool for measuring the association between a possible cause X and a particular effect Y. The elements of the target population are divided in two groups according to whether X is present or not, the absence of X identifies the control group. A random sample is selected from each group with the aim of measuring the association between X and Y. We propose an unbiased estimator of the associational risk difference between groups and then generalize our approach to the problem of estimating the causal risk difference.

"\"In this paper we propose an extended version of a model-based strategy for clustering spatial functional data. The strategy, we refer, aims simultaneously to classify spatially dependent curves and to obtain a spatial functional model prototype for each cluster. The fit of these models implies to estimate a variogram function, the trace variogram function. Our proposal is to introduce an alternative estimator for the trace-variogram function: a kernel variogram estimator. This works better to adapt spatial varying features of the functional data pattern. Experimental comparisons show this approach has some advantages over the previous one. \""

In this paper, a new approach to the fuzzy analysis of multidimensional material deprivation data is provided, based on partial order theory. The main feature of the methodology is that the information needed for the deprivation assessment is extracted directly from the relational structure of the dataset, avoiding any kind of scaling and aggregation procedure, so as to respect the ordinal nature of the data. An example based on real data is worked out, pertaining to material deprivation in Italy for the year 2004.

This paper extends the scedasticity comparison among several groups of observations, usually complying with the homoscedastic and the heteroscedastic cases, in order to deal with data sets laying in an intermediate situation. As is well known, homoscedasticity corresponds to equality in orientation, shape and size of the group scatters. Here our attention is focused on two weaker requirements: scatters with the same orientation, but with different shape and size, or scatters with the same shape and size but different orientation. We introduce a multiple testing procedure that takes into account each of the above conditions. This approach discloses a richer information on the data underlying structure than the classical method only based on homo/heteroscedasticity. At the same time, it allows a more parsimonious parametrization, whenever the patterned model is appropriate to describe the real data. The new inferential methodology is then applied to some well-known data sets, chosen in the multivariate literature, to show the real gain in using this more informative approach. Finally, a wide simulation study illustrates and compares the performance of the proposal using data sets with gradual departure from homoscedasticity.

We consider constrained formulations of the maximum likelihood estimation for mixture models. In this framework, we introduce the concept of weak homoscedasticity for covariance matrices of the component densities and give a test for detecting weak homoscedasticity in two sample data under the multinormal assumption. Based on such concept, we present a constrained EM algorithm for data modeling via mixtures of t-distributions. The proposal is illustrated on the ground of numerical experiments which show the usefulness of the present approach in data modeling.