Mini-course: Advanced Stationary Processes Analysis (Time Series Analysis)

Thời gian: 09:00:14/07/2016 đến 17:00:22/07/2016

Địa điểm: B4-705

Báo cáo viên: Jean-Yves Dauxois, INSA-IMT University of Toulouse, France; Vincent Lefieux, RTE Paris, France.

Tóm tắt:

The time series analysis is a very well known part of Statistics and used in many areas like: economics, finance, environment, energy, medicine, geophysics... The analysis of such stochastic processes is strongly based on the property of stationarity. The aim of this lecture is to introduce some extensions of this models and to study their statistical inference.

One part of the lecture will be devoted to multivariate time series and an introduction of spatial time series. In dealing with economic time series, we often need to forecast stochastic processes based non only on their own past values but also on other processes. In order to achieve this, we can use VAR models, cointegration theory and state space models which constitute a nice statistical framework to address common challenges.

Another part of the lectures will consider the field of Geostatistics which aim is to make statistical inference on some kinds of Spatial processes. Assuming a weaker notion of Stationarity (Intrinsic property), one can estimate a variable of interest over a whole domain (seen as a realization of a spatial process) on the basis of the observation on a limited number of points. The Kriging is the most used technique in this area. Applications are often encountered in hydrology, meteorology, oceanography, geography, among others...

Lectures will be accompanied by tutorials and computer lab works with R software.

Ngôn ngữ sử dụng trong bài giảng: Tiếng Anh.

Download Lecture:

1. Mini-course-Dauxois.pdf.

2. Mini-course-Dauxois-2.pdf

3. TimesSeriesIntroduction.R

4. TP_Geostat_Rmd.html