HYBRID CONFERENCE

New Results on Time Series and their Statistical Applications
Séries chronologiques: nouveaux résultats et applications statistiques

14 au 18 septembre 2020

Scientific Committee
Comité scientifique

Paul Doukhan (​Université de Cergy-Pontoise)
Liudas Giraitis (Queen Mary University of London)
Suhasini Subba Rao (Texas A&M University)
Olivier Winterberger (Sorbonne Université)

Organizing Committee
Comité d’organisation

Jean-Marc Bardet (Université Paris 1 Panthéon-Sorbonne)
​Idris Eckley (Lancaster University)
Konstantinos Fokianos (University of Cyprus)
Michael H. Neumann (Friedrich Schiller University Jena)
Anne Philippe (Université de Nantes)

Description
Statistical modeling and analysis of time series data has been traditionally based on the assumption of stationarity and/or low dimensionality. Nowadays technology calls for abandoning such assumptions and require the development of new and more sophisticated statistical methods. For instance, the assumption of stationarity, which has dominated most of time series literature, is restricted and cannot be justified for many applications. Additionally, study of long-range dependent processes revealed concrete distinctions between stationary and non-stationary processes. And it is quite common in practice to observe time series which exhibit trend, periodic behavior and require additionally inclusion of covariates.

These new applications are developing very quickly and are found in diverse research areas. Based on this fact and motivated by these challenging problems, we propose to organize a conference whose main theme is New results on time series and their statistical applications. We are quite confident that such a meeting will bring together researchers from all over the world to discuss and exchange ideas about future research directions and initiatives.

In particular, we aim on a program that covers many ”hot” research topics like:
​• Locally-stationary and non-stationary time series;
• High-dimensional time series;
• Change-point methods and switching processes;
• Long-range dependent time series;
• New limit theorems on means and extremes;
• New concentration inequalities and model selection;
• New measures of dependence;
• New statistical applications on reall datasets

La modélisation statistique et l’analyse des données de séries chronologiques ont été traditionnellement basées sur l’hypothèse de stationnarité et/ou de faible dimensionnalité. De nos jours, la technologie exige l’abandon de ces hypothèses et le développement de méthodes statistiques nouvelles et plus sophistiquées. Par exemple, l’hypothèse de stationnarité, qui a été retenue et a dominé la plupart de la littérature sur les séries chronologiques, est restreinte et ne peut être justifiée pour de nombreuses applications. Par ailleurs, l’étude des processus à longue mémoire a permis d’affiner les distinctions entre processus stationnaires et non stationnaires. Et c’est assez courant dans la pratique d’observer des séries temporelles qui présentent des tendances, des comportements périodiques et nécessitant l’inclusion de covariables.

Ces nouvelles directions se développent très rapidement et se retrouvent dans divers domaines de recherche. Partant de ce constat et motivés par ces problèmes difficiles, nous proposons d’organiser une conférence dont le thème principal est Séries chronologiques: nouveaux résultats et applications statistiques. Nous sommes convaincus qu’une telle réunion rassemblera des chercheurs du monde entier pour discuter des questions suivantes et échanger des idées sur les orientations et les initiatives de recherche futures.

En particulier, nous visons un programme qui couvrira de nombreux sujets de recherche comme:
• Séries temporelles localement stationnaires et non stationnaires;
• Séries temporelles de haute dimension;
• Détections de ruptures et processus àchangement de régime;
• Séries temporelles fortement dépendantes;
• Nouveaux théorèmes limite;
• Nouvelles inégalités de concentration et sélection de modèles;
• Nouvelles mesures de dépendance;
​• Nouvelles applications statistiques sur données réelles.

 Speakers

Alexander Aue (University of California, Davis)    Relevant two-sample tests for the eigenfunctions of covariance operators
Felix Cheysson (AgroParis Tech)    Estimation of Hawkes processes from binned observations using Whittle likelihood
Richard Davis (Columbia University)    Modeling of Time Series Using Random Forests: Theoretical Developments
Herold Dehling (Ruhr University Bochum)    An Asymptotic Test for Constancy of Variance under Short-Range Dependence
Holger Dette (Ruhr University Bochum)     Relevant hypotheses in functional data
Youssef Esstafa (Université de Franche-Comté)    Estimating FARIMA models with uncorrelated but non-independent error terms
Christian Francq (CREST)     Adaptiveness of the empirical distribution of residuals in semi-parametric conditional location scale models
Liudas Giraitis (Queen Mary University of London)    Time-Varying Instrumental Variable Estimation
Yannig Goude (EDF R&D, Université Paris-Saclay)    Machine learning methods for electricity load forecasting: contributions and perspectives
Kamila Kare (Université Panthéon Sorbonne)     Consistent model selection criteria and goodness-of-fit test for common time series models
Claudia Kirch (Otto von Guericke University of Magdeburg)    Beyond Whittle’s likelihood: New Bayesian semiparametric approaches to time series analysis
Claudia Klueppelberg (Technical University of Munich)    Indirect Inference for Time Series Using the Empirical Characteristic Function and Control Variates
Clifford Lam (The London School of Economics and Political Science)    Robust mean and Eigenvalues regularized covariance matrix estimation
Émilie Lebarbier (Université Paris Nanterre)     A factor model approach for the joint segmentation of correlated series
Kathryn Leeming (University of Warwick)    Modelling farm temperatures with irregular seasonality
Remigijus Leipus (Vilnius University)     Estimating and testing long memory in random coefficient dynamic panel data model
Anne Leucht (University of Bamberg)    Testing Equality of Spectral Density Operators for Functional Processes
Philippe Naveau (CNRS, Université Versailles Saint Quentin)    Detecting seasonality changes in multivariate extremes from climatological time series
Riccardo Passeggeri  (Sorbonne Université)     Asymptotic analysis of extremes of general stationary spatio-temporal models
Joseph Rynkiewicz (Université Panthéon Sorbonne)     Mixtures of Nonlinear Poisson Autoregressions
Philippe Soulier (Université Paris Nanterre)    The tail process and tail measure of continuous time regularly varying stochastic processes
Suhasini Subba Rao (Texas A&M University)     Reconciling the Gaussian and Whittle Likelihood with an application to estimation in the frequency domain
Lionel Truquet (ENSAI)    Stationarity and ergodic properties for some observation-driven models in random environment
Almut Veraart (Imperial College London)     Likelihood-based estimation, model selection, and forecasting of integer-valued trawl processes
Rainer von Sachs (Université Catholique de Louvain)    Intrinsic wavelet smoothing of curves of Hermitian positive definite matrices (with applications to spectral density estimation of multivariate time series)
Jean-Michel Zakoian (CREST)     Testing the existence of moments for GARCH-type processes
Zhou Zhou (University of Toronto)    Frequency Detection and Change Point Estimation for Time Series of Complex Oscillation
SPONSORS