Model-Based Clustering and Classification for Data Science: With Applications in R

Description

Cluster analysis finds groups in data automatically. Most methods have been heuristic and leave open such central questions as: how many clusters are there? Which method should I use? How should I handle outliers? Classification assigns new observations to groups given previously classified observations, and also has open questions about parameter tuning, robustness and uncertainty assessment. This book frames cluster analysis and classification in terms of statistical models, thus yielding principled estimation, testing and prediction methods, and sound answers to the central questions. It builds the basic ideas in an accessible but rigorous way, with extensive data examples and R code; describes modern approaches to high-dimensional data and networks; and explains such recent advances as Bayesian regularization, non-Gaussian model-based clustering, cluster merging, variable selection, semi-supervised and robust classification, clustering of functional data, text and images, and co-clustering. Written for advanced undergraduates in data science, as well as researchers and practitioners, it assumes basic knowledge of multivariate calculus, linear algebra, probability and statistics.

Charles Bouveyron is Full Professor of Statistics at Universite Cote d'Azur and the Chair of Excellence in Data Science at Institut National de Recherche en Informatique et en Automatique (INRIA), Rocquencourt. He has published extensively on model-based clustering, particularly for networks and high-dimensional data. Gilles Celeux is Director of Research Emeritus at Institut National de Recherche en Informatique et en Automatique (INRIA), Rocquencourt. He is one of the founding researchers in model-based clustering, having published extensively in the area for thrity-five years. T. Brendan Murphy is Full Professor in the School of Mathematics and Statistics at University College Dublin. His research interests include model-based clustering, classification, network modeling and latent variable modeling. Adrian E. Raftery is the Boeing International Professor of Statistics and Sociology at the University of Washington. He is one of the founding researchers in model-based clustering, having published in the area since 1984.

'Bouveyron, Celeux, Murphy, and Raftery pioneered the theory, computation, and application of modern model-based clustering and discriminant analysis. Here they have produced an exhaustive yet accessible text, covering both the field's state of the art as well as its intellectual development. The authors develop a unified vision of cluster analysis, rooted in the theory and computation of mixture models. Embedded R code points the way for applied readers, while graphical displays develop intuition about both model construction and the critical but often-neglected estimation process. Building on a series of running examples, the authors gradually and methodically extend their core insights into a variety of exciting data structures, including networks and functional data. This text will serve as a backbone for graduate study as well as an important reference for applied data scientists interested in working with cutting-edge tools in semi- and unsupervised machine learning.' John S. Ahlquist, University of California, San Diego 'This book, written by authoritative experts in the field, gives a comprehensive and thorough introduction to model-based clustering and classification. The authors not only explain the statistical theory and methods, but also provide hands-on applications illustrating their use with the open-source statistical software R. The book also covers recent advances made for specific data structures (e.g. network data) or modeling strategies (e.g. variable selection techniques), making it a fantastic resource as an overview of the state of the field today.' Bettina Grun, Johannes Kepler Universitat Linz, Austria 'Four authors with diverse strengths nicely integrate their specialties to illustrate how clustering and classification methods are implemented in a wide selection of real-world applications. Their inclusion of how to use available software is an added benefit for students. The book covers foundations, challenging aspects, and some essential details of applications of clustering and classification. It is a fun and informative read!' Naisyin Wang, University of Michigan 'This is a beautifully written book on a topic of fundamental importance in modern statistical science, by some of the leading researchers in the field. It is particularly effective in being an applied presentation - the reader will learn how to work with real data and at the same time clearly presenting the underlying statistical thinking. Fundamental statistical issues like model and variable selection are clearly covered as well as crucial issues in applied work such as outliers and ordinal data. The R code and graphics are particularly effective. The R code is there so you know how to do things, but it is presented in a way that does not disrupt the underlying narrative. This is not easy to do. The graphics are 'sophisticatedly simple' in that they convey complex messages without being too complex. For me, this is a 'must have' book.' Rob McCulloch, Arizona State University 'This advanced text explains the underlying concepts clearly and is strong on theory ... I congratulate the authors on the theoretical aspects of their book, it's a fine achievement.' Antony Unwin, International Statistical Review 'In my opinion, the overall quality of this impactful and intriguing book can be expressed by concluding that it is a perfect fit to the Cambridge Series in Statistical and Probabilistic Mathematics, characterized as a series of high-quality upper-division textbooks and expository monographs containing applications and discussions of new techniques while emphasizing rigorous treatment of theoretical methods.' Zdenek Hlavka, MathSciNet '... this book not only gives the big picture of the analysis of clustering and classification but also explains recent methodological advances. Extensive real-world data examples and R code for many methods are also well summarized. This book is highly recommended to students in data science, as well as researchers and data analysts.' Li-Pang Chen, Biometrical Journal 'Model-Based Clustering and Classification for Data Science: With Applications in R, written by leading statisticians in the field, provides academics and practitioners with a solid theoretical and practical foundation on the use of model-based clustering methods ... this book will serve as an excellent resource for quantitative practitioners and theoreticians seeking to learn the current state of the field.' C. M. Foley, Quarterly Review of Biology 'This book frames cluster analysis and classification in terms of statistical models, thus yielding principled estimation, testing and prediction methods, and sound answers to the central questions ... Written for advanced undergraduates in data science, as well as researchers and practitioners, it assumes basic knowledge of multivariate calculus, linear algebra, probability and statistics.' Hans-Jurgen Schmidt, zbMATH