The workshop will encompass recent research works on Time Series and Extremes.
A non exhaustive list of subjects covered is :
- time series modelling including long memory, non-causality or discrete value effects ;
- serially dependent extremes and regularly varying time series.
A short course on regularly varying time series will be given by Philippe Soulier.
PROGRAM
Thursday May 16, 2019
- 09:30 : Welcome & Coffee
- 10:00 : Philippe SOULIER (Université Paris Nanterre), Heavy tailed time series : models, limit theorems, estimation of extremal characteristics
- 11:30 : Christian FRANCQ (Université Lille 3 et CREST), Observation-driven count and duration time series models
- 12:30 : Lunch
- 14:00 : François ROUEFF (Telecom ParisTech), About linear and duration-driven long range dependence
- 15:00 : Jean-Marc BARDET (Université Paris 1 Panthéon-Sorbonne), Local semi-parametric estimation for locally stationary process with infinite memory
- 16:00 : Coffee break
- 16:30 : Olivier WINTENBERGER (Sorbonne Université), Extreme Value Theory for the Diagonal BEKK-ARCH(1) Model
Friday May 17, 2019
- 09:00 : Philippe SOULIER (Université Paris Nanterre), Heavy tailed time series : models, limit theorems, estimation of extremal characteristics
- 10:30 : Coffe break
- 11:00 : Holger DREES (University of Hambourg), Block based extreme value statistics : disjoint vs moving blocks
- 12:00 : Anja JANSSEN (Royal Institute of Technology in Stockholm), Dimension and complexity reduction for multivariate and serially dependent extremes
- 13:00 : Lunch
- 14:00 : Anne PHILIPPE (Université de Nantes), Testing for long memory in panel random-coefficient AR(1) data
- 15:00 : Jean-Michel ZAKOIAN (CREST), Mixed Causal-Noncausal AR process and the modeling of bubbles
► Organizers : Yacouba BOUBACAR MAINASSARA and Clément DOMBRY
► Information booklet (detailed program, abstracts)