Week 3: Process
February 15-19, 2016
One of the features present in the actual data from many areas (Economy, Finance, Hydrology, Astrophysics) is the presence of dependency. This dependence between observations may manifest in several degrees (low dependency, high dependency).
Statistical inference for such data requires a more suitable theorems of more suitable central limit theorems.
The time series analysis is often based on the assumption of stationarity. However the actual data observed in many areas are often non-stationary.
Examples include, stock market cracks often induce breaks in macroeconomic and financial time series, climate change involves changing settings in the generating process of the temperature data, hence the need to introduce more suitable statistical models (locally stationary processes, processes with breaks).

The aim of this international meeting, declined in two themes, is:

– Make a state of the art of different scientific advances in each of the two themes
– Open new directions for research and opportunities for young researchers

Scientific Committee

Jean-Marc Bardet (Université Paris 1- Panthéon-Sorbonne)
Rainer Dahlhaus (University of Heidelberg)
Paul Doukhan (Université de Cergy Pontoise)

Organizing Committee

Mohamed Boutahar (Aix-Marseille Université)
Laurence Reboul (Aix-Marseille Université)


Stationary increments harmonizable stable fields: upper estimates on path behavior

Asymptotic behavior of the Laplacian quasi-maximum likelihood estimator of affine causal processes

Verfitting of the Hurst index for a multifractional Brownian motionr

Large scale reduction principle and application to hypothesis testing

Behavior of the Wasserstein distance between
the empirical and the marginal distributions of alpha-dependent sequences

Detecting long-range dependence in non-stationary time series

Semi-parametric dynamic factor models for non-stationary time series

Mallows’ Quasi-Likelihood Estimation for Log-linear Poisson

Hölderian weak invariance principle for strictly stationary sequences 

Quasi-MLE for quadratic ARCH model with long memory

Statistical inference for bifurcating Markov chains

Phantom distribution functions for dependent random vectors

Testing for parameter change in a general class of time seriesof counts

Segmentation of time-series with various types of dependency

Processes with varying local regularities

Strong approximation for additive functionals of geometrically ergodic Markov chains

A test for local white noise (and the absence of aliasing) in locallystationary wavelet time series

Periodogram Based Tests of Stationarity

Adaptive bandwidth selection with cross validation for locallystationary processes

Posterior consistency for partially-observed Markov models

Fonctional limit theorems for weakly dependent regularly vary-ing time series

Fourier based statistics for irregular spaced spatial data

Detecting a changed segment in a sample

Some further properties and applications  of local Gaussian approximation

Parameter stability and semiparametric inference in time-varyingARCH processes

Classification of nonparametric time trends

Martingale central limit theorems for random fields

Time-frequency analysis of locally stationary Hawkes processes