The European Meeting of Statisticians (EMS), sponsored by the European Regional Committee of the Bernoulli Society, is the main conference in statistics and probability in Europe. EMS is a conference where statisticians of all ages and from all regions meet to exchange ideas and talk about the newest developments on the broad field of statistics and probability theory. The conference is organized every other year. The very first EMS meeting was held in Dublin in 1962, and the 32nd EMS - EMS 2019 - will take place in Palermo, the capital of Sicily (Italy).
Palermo is a city on the northwestern coast of Sicily, on the Tyrrhenian Sea. It is a city rich in monuments that testify its centuries-old history. Some monuments are of particular interest: the Norman Palace (Palazzo dei Normanni), built around 1100, which is the oldest Royal residence in Europe, inside the Palace there is also the famous Cappella Palatina; the Cathedral, built in its present form in 1185 on the basis of an older church dated from the 4th century; the Martorana Church (Chiesa della Martorana) of Arabic style built in 1143; the Teatro Massimo, built in 1875, which is the largest opera house in Italy and one of the largest in Europe. Very close to Palermo is the village of Mondello where there is one of the most beautiful beach in Sicily. Palermo is also famous for its traditional food. There are also many good restaurants for vegan and vegetarian. Palermo is well connected by plane with the main European cities and there are many flight connections with Rome and Milan. From the airport is very easy to reach the city by train (there is a train station inside the airport), by bus (there are buses from the airport to the city and vice versa every half an hour) and by taxi. There are also different accommodation possibilities in Palermo, in different price ranges.
The conference will be held at the main campus of the University of Palermo , which is located in the South part of the city, but very close to the city centre of Palermo. The conference will start on Monday 22.7.2019 and it will end on Friday 26.7.2019.   The program consists of invited and contributed lectures, and posters, addressing a full range of subjects in statistics and its many applications.


Plenary Lectures

> <
  • Judith Rousseau, Oxford University:

    Bayesian measures of uncertainty

    Abstract: The reknown theorem of Bernstein von Mises in regular finite dimensional models has numerous interesting consequences, in particular it implies that a large class of credible regions are also asymptotically confidence regions, which in turns imply that different priors lead to the same credible regions to first order. Unfortunately the Bernstein von Mises theorem does not necessarily hold in high or infinite dimensional models and understanding the asymptotic behaviour of credible regions is much more involved. In this talk I will describe what are the new advances that have been obtained over the last 8 years or so on the understanding - or not- of credible regions in semi and non- parametric models. I will in particular discuss some interesting phenomena which have been exhibited in high dimensional models, for certain families of priors, encountered for instance in mixture models with unknown number of components, in regression models with a large number of covariates etc... We can show that in a significant number of cases these priors tend to over penalize (or over smooth), leading to only partially robust confidence statements. I will also discuss the few advances which have been obtained in the context of non or semi parametric mixture models, which are notoriously difficult to study.

  • Genevera Allen, Rice University:

    Data Integration: Data-Driven Discovery from Diverse Data Sources.

    Abstract: Data integration, or the strategic analysis of multiple sources of data simultaneously, can often lead to discoveries that may be hidden in individual analyses of a single data source. In this talk, we present several new techniques for data integration of mixed, multi-view data where multiple sets of features, possibly each of a different domain, are measured for the same set of samples. This type of data is common in heathcare, biomedicine, national security, multi-senor recordings, multi-modal imaging, and online advertising, among others. In this talk, we specifically highlight how mixed graphical models and new feature selection techniques for mixed, mutli-view data allow us to explore relationships amongst features from different domains. Next, we present new frameworks for integrated principal components analysis and integrated generalized convex clustering that leverage diverse data sources to discover joint patterns amongst the samples. We apply these techniques to integrative genomic studies in cancer and neurodegenerative diseases to make scientific discoveries that would not be possible from analysis of a single genomics data set.

  • Aad Van Der Vaart, University of Leiden:

    Nonparametric Bayes: review and challenges.

    Abstract: Nonparametric Bayesian methods have seen a great development in the past decades. They are ordinary Bayesian methods using a prior distribution on an infinite-dimensional or high-dimensional parameter (function, distribution, high-dimensional regression vector), resulting in a posterior distribution, giving the plausibility of this parameter given the data. Nonparametric Bayesian methods are now routinely applied in many areas, based on the promise of reconstruction, through the mode or mean of the posterior distribution, and automatic uncertainty quantification, through the spread in the posterior distribution. Besides from statisticians they attract attention of computer scientists and mathematical analysts, in particular in connection with inverse problems and data assimilation. Through empirical Bayes ideas they are connected to the `sharing of information' and large scale inference in settings of high-dimensional data. There is an increasing theoretical understanding of the performance of these methods from the non-Bayesian perspective, developed under the assumption that the prior is only a working hypothesis to model a true state of nature. Theory has been developed for classical nonparametric smoothing problems, sparse high-dimensional models and increasingly for inverse problems, and addresses a great variety of priors, based on the classical Dirichlet process, Gaussian processes, spike-and-slab distributions, and many others. One of the attractions is the automatic adaption to complexity by means of hyperparameters fitted through hierarchical and empirical Bayes approaches. Theory addresses rates of contraction of the posterior to a true parameter, distributional approximations of the posterior distribution of smooth functionals, and most recently the coverage (or lack of it) of Bayesian credible sets. In this talk we present some examples of success stories, and point to open questions.

  • Victor M. Panaretos, Ecole Polytechnique Federale de Lausanne:

    Amplitude and Phase Variation of Random Processes.

    Abstract: The amplitude variation of a random process consists of random oscillations in its range space (the ``y-axis''), typically encapsulated by its (co)variation around a mean level. In contrast, phase variation refers to fluctuations in its domain (the ``x-axis''), often caused by random time changes or spatial deformations. Many types of processes, particularly physiological processes, manifest both types of variation, and confounding them can seriously skew statistical inferences. We will consider some of the statistical challenges related to empirically separating these two forms of variation, a problem also known as registration, synchronisation, or multireference alignment, in other contexts. Our approach will largely rely on the tools and geometry of optimal (multi)transport, and borrow from connections to notions from shape theory, such as Procrustes analysis and tangent space PCA. The approach will hopefully also highlight the intriguing aspect of this problem, as being at the confluence of functional data analysis, where the data are elements of infinite dimensional vector spaces, and geometrical statistics, where the data are elements of differentiable manifolds.

  • Gilles Blanchard, University of Potsdam:

    Sketched learning using random moments. (with R. Gribonval, N. Keriven and Y. Traonmilin).

    Abstract: We introduce and analyze a general framework for resource-efficient large-scale statistical learning by data sketching: a training data collection is compressed in one pass into a low-dimensional sketch (a vector of random empirical generalized moments) that should capture the information relevant to the considered estimation task. The estimation target is the minimizer of the population risk for a given loss function. An approximate minimizer of the empirical risk is computed from the sketch information only using a constrained moment matching principle. Sufficient sketch sizes to control the statistical error of this procedure are investigated. This principle is applied to different setups: PCA, clustering, and Gaussian mixture Modeling.

  • John Lafferty, Yale University:

    Computational perspectives on some statistical problems.

    Abstract: We present some variations on classical statistical problems that take a computational, machine learning perspective. First, we study nonparametric regression when the data are distributed across multiple machines. We place limits on the number of bits that each machine can use to transmit information to the central machine. Our results give both asymptotic lower bounds and matching upper bounds on the statistical risk under various settings. Second, we investigate the use of machine learning algorithms that are required to obey natural shape constraints suggested by domain knowledge. We develop methods for high-dimensional shape-constrained regression and classification that "reshape" the original prediction rule. Finally, we study optimization procedures to minimize the empirical risk functional for certain families of deep neural networks. We develop an approach that optimizes a sequence of objective functions using network parameters obtained during different stages of the learning process. This is evaluated with deep generative networks used as a replacement for sparsity in compressed sensing and approximation.