Extreme Value Theory in R
Synopsis
The Extreme Value Theory (EVT) is extensively used for modelling very large and/or very small events. Usually the focus of the analysis is the estimation of very extreme quantiles or tail probabilities. It is widely used in several areas, such as environment, insurance, whether and hydrology.
Several R packages have been developed for fitting models in this framework.
This course aims to present the most common statistical tools of the EVT, as well as teaching the use of a R package.
Several R packages have been developed for fitting models in this framework.
This course aims to present the most common statistical tools of the EVT, as well as teaching the use of a R package.
Target audience
Users and researchers from the Environment/Biology areas.
Topics
 Why do we need the Extreme Value Theory (EVT)?
 The purpose of the EVT
 The Classical EVT  The extreme value distribution (GEV)
 The Peaks Over Threshold (POT)  The generalized Pareto distribution (GPD)
 Descriptive ways of assessing tail heaviness
 The extreme quantile estimation
 Methods for estimating the parameters of the GEV and the GPD
 R packages for performing an EVT analysis
 Applications (Practical sessions)
Schedule
This is a twodays course with the following structure:
 09:30  11:00 Lecture (computer lab)
 11:00  11:30 Coffee break
 11:30  13:00 Lecture (computer lab)
 13:00  14:00 Lunch (not provided)
 14:00  15:30 Lecture (computer lab)
 15:30 16:00 Coffee break
 16:00  17:30 Lecture (computer lab)
Lecturer
Patrícia Zea Bermudez, Professor at the Departament of Statistics and Operational Research in the University of Lisbon and full member of the Center of Statistics and Applications of University of Lisbon
Registration

