Skip to main content

Statistics Seminar by Hernando Ombao & Philippe Naveau

isba
    • 11 Oct
  • Accessible
More information

11:00 : Hernando Ombao (C.035)
14:30 : Philippe Naveau (C.115)

11:00 : Hernando Ombao (C.035)

King Abdullah University of Science and Technology (KAUST) - Saudi Arabia
Invited by Rainer von Sachs

Connectivity in a Brain Network: An Overview

Abstract:

The brain can be considered as a system consisting of nodes at different levels starting from the microscopic to macroscopic: neurons, local fields, anatomical voxels, channels, regions of interest. One of the key problems in studying brain function is connectivity which reflects how different nodes of the brain interact with each other during rest and during a cognitive task. The first part of the talk will cover background material on the measures of functional connectivity which covers both time domain (cross-correlation) and frequency domain (coherence, partial coherence). Here, we also present our proposed metric, which we call spectral transfer entropy (STE) which quantifies the magnitude and direction of information flow from a certain frequency-band oscillation of a channel to an oscillation of another channel. Our approach utilizes the vine copula representation under which it is straightforward to test for the null hypothesis (zero STC) through a standard resampling approach.
In the second part of the talk, we consider the problem of characterizing "global” dependence between two regions, each consisting of many nodes. The naïve approach of computing all pairwise dependence is computationally demanding. Our approach is inspired by canonical coherence analysis (CCA) and in essence examines summarized dependence between the oscillatory activity at the pair of regions.  To overcome the limitations of CCA (linearity and sensitivity to outliers), we use the Kendall’s tau as a measure of dependence between the summarized oscillatory activity.
These proposed spectral dependence measures will be used to examine human electeoencephalograms (EEGs) and calcium recordings in mice.
This talk is based on joint work with Paolo Redondo, Mara Talento, and Sarbojit Roy.

---

14:30 : Philippe Naveau (C.115)

Laboratoire des Sciences du Climat et de l'Environnement (LSCE)
Invited by Eugene PIRCALABELU

will give a presentation on :

Statistical modelling of records in a changing climate

Abstract:

The increase of  recent climate record frequencies raises many statistical and climatological questions. Statistically,  the literature on modelling  record changes in a non-stationary context is sparse and, in this context,  I will present different case studies that blind multivariate extreme value theory and counterfactual theory.  Another aspect will be to determine how this non-stationary allows to improve the forecasting of record frequencies. If time allowed, the choice of climatological explanatory variables (co-variates) to understand unprecedented heatwaves will be touched upon. All statistical techniques will be illustrated by cases studies based either on climate model outputs or weather stations recordings. Concerning the temporal periods, these examples will treat past, present and future yearly time scales.

  • Friday, 11 October 2024, 08h00
    Friday, 11 October 2024, 17h00
  • Contact