1,720,981 research outputs found
Model-based clustering with sparse covariance matrices
No full textFinite Gaussian mixture models are widely used for model-based clustering of continuous data. Nevertheless, since the number of model parameters scales quadratically with the number of variables, these models can be easily over-parameterized. For this reason, parsimonious models have been developed via covariance matrix decompositions or assuming local independence. However, these remedies do not allow for direct estimation of sparse covariance matrices nor do they take into account that the structure of association among the variables can vary from one cluster to the other. To this end, we introduce mixtures of Gaussian covariance graph models for model-based clustering with sparse covariance matrices. A penalized likelihood approach is employed for estimation and a general penalty term on the graph configurations can be used to induce different levels of sparsity and incorporate prior knowledge. Model estimation is carried out using a structural-EM algorithm for parameters and graph structure estimation, where two alternative strategies based on a genetic algorithm and an efficient stepwise search are proposed for inference. With this approach, sparse component covariance matrices are directly obtained. The framework results in a parsimonious model-based clustering of the data via a flexible model for the within-group joint distribution of the variables. Extensive simulated data experiments and application to illustrative datasets show that the method attains good classification performance and model quality. The general methodology for model-based clustering with sparse covariance matrices is implemented in the R package mixggm, available on CRAN
CLADAG 2019 Special Issue: Selected Papers on Classification and Data Analysis (editoriale)
No full textThis special issue of Statistical Analysis and Data Mining collects papers presented at the 12-th Scientific Meeting of the Classification and Data Analysis Group (CLADAG) of the Italian Statistical Society (SIS), held in Cassino, Italy, September 11 – 13, 2019.
The CLADAG group, founded in 1997, promotes advanced methodological research in multivariate statistics with a special vocation in Data Analysis and Classification. CLADAG is a member of the International Federation of Classification Societies (IFCS). It organizes a biennial international scientific meeting, schools related to classification and data analysis, publishes a newsletter, and cooperates with other member societies of the IFCS to the organization of their conferences. Founded in 1985, the IFCS is a federation of national, regional, and linguistically-based classification societies aimed at promoting classification research.
Previous CLADAG meetings were held in Pescara (1997), Roma (1999), Palermo (2001), Bologna (2003), Parma (2005), Macerata (2007), Catania (2009), Pavia (2011), Modena and Reggio Emilia (2013), Cagliari (2015), and Milano (2017).
Best papers from the conference have been submitted to this special issue, and five of them have been selected for publication, following a blind peer-review process. The manuscripts deal with different data analysis issues: mixture of distributions, compositional data analysis, Markov chain for web usability, survival analysis, and applications to high-throughput, eye-tracking, and insurance transaction data.
The paper by S.X. Lee et al. proposes a parallelization strategy of the Expectation-Maximization (EM) algorithm, with a special focus on the estimation of finite mixtures of flexible distribution such as the canonical fundamental skew t distribution (CFUST). The parallel implementation of the EM-algorithm is suitable for single-threaded and multi-threaded processors as well as for single machine and multiple-node systems.
The EM algorithm is also discussed in the paper of L. Scrucca. Here, a fast and efficient Modal EM algorithm for identifying the modes of a density estimated through a finite mixture of Gaussian distributions with parsimonious component covariance structures is provided. The proposed approach is based on an iterative procedure aimed at identifying the local maxima, exploiting features of the underlying Gaussian mixture model.
Motivated by applications in high-throughput compositional data analysis, the paper by N. Štefelová et al. proposes a data-driven weighting strategy to enhance marker identification through PLS regression with compositional predictors. The weighting strategy draws on the correlation structure between response variable and pairwise log-ratios. Its practical relevance is illustrated through an analysis of metabolite signals associated with the emission of greenhouse gases from cattle.
The paper by G. Zammarchi et al. exploits Markov chain to analyse web usability of a University website using eye tracking methodology. With the aim of improving its usability, the paper compares performances of high school and University students in terms of time to completion, number of fixations and difficulty ratio across ten different tasks.
Data from a commercial insurance company in the Czech Republic are instead exploited by D. Zapletal to compare the efficacy of some survival analysis models within an insurance transaction framework. The ability to identify relevant explanatory variables through the Cox proportional hazard model and some competing risk models (i.e., the cause-specific and the sub-distribution hazard models) is assessed on a large data set consisting of more than 200 thousand individuals.
In brief, this special issue is in line with the CLADAG goal of supporting the interchange of ideas in Classification and Data Analysis. We strongly believe it well represents the scientific characteristics of the CLADAG community, and we invite all readers to join the next CLADAG conference, which will be in Florence, September 11 to 13, 2021
Mclust 5: Clustering, classification and density estimation using Gaussian finite mixture models
No full textFinite mixture models are being used increasingly to model a wide variety of random phenomena for clustering, classification and density estimation. mclust is a powerful and popular package which allows modelling of data as a Gaussian finite mixture with different covariance structures and different numbers of mixture components, for a variety of purposes of analysis. Recently, version 5 of the package has been made available on CRAN. This updated version adds new covariance structures, dimension reduction capabilities for visualisation, model selection criteria, initialisation strategies for the EM algorithm, and bootstrap-based inference, making it a full-featured R package for data analysis via finite mixture modelling
Model-Based Clustering, Classification, and Density Estimation Using mclust in R
No full textModel-based clustering and classification methods provide a systematic statistical approach to clustering, classification, and density estimation via mixture modeling. The model-based framework allows the problems of choosing or developing an appropriate clustering or classification method to be understood within the context of statistical modeling. The mclust package for the statistical environment R is a widely adopted platform implementing these model-based strategies. The package includes both summary and visual functionality, complementing procedures for estimating and choosing models. Key features of the book: An introduction to the model-based approach and the mclust R package A detailed description of mclust and the underlying modeling strategies An extensive set of examples, color plots, and figures along with the R code for reproducing them Supported by a companion website, including the R code to reproduce the examples and figures presented in the book, errata, and other supplementary material Model-Based Clustering, Classification, and Density Estimation Using mclust in R is accessible to quantitatively trained students and researchers with a basic understanding of statistical methods, including inference and computing. In addition to serving as a reference manual for mclust, the book will be particularly useful to those wishing to employ these model-based techniques in research or applications in statistics, data science, clinical research, social science, and many other disciplines
Robust variable selection for model-based learning in presence of adulteration
Full text linkThe problem of identifying the most discriminating features when performing supervised learning has been extensively investigated. In particular, several methods for variable selection have been proposed in model-based classification. The impact of outliers and wrongly labeled units on the determination of relevant predictors has instead received far less attention, with almost no dedicated methodologies available. Two robust variable selection approaches are introduced: one that embeds a robust classifier within a greedy-forward selection procedure and the other based on the theory of maximum likelihood estimation and irrelevance. The former recasts the feature identification as a model selection problem, while the latter regards the relevant subset as a model parameter to be estimated. The benefits of the proposed methods, in contrast with non-robust solutions, are assessed via an experiment on synthetic data. An application to a high-dimensional classification problem of contaminated spectroscopic data is presented
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
- …
