Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
This lecture on Vector Autoregression covers the modeling of vector-valued time series using autoregressive processes. Starting from AR processes, the instructor explains the generalization to vector-valued processes and the concept of Vector Auto-Regression. The lecture delves into the stationarity of VAR models, including the conditions for stability. Examples of VAR(1) processes are provided to illustrate stability analysis. The lecture also discusses the Yule-Walker equations for VAR processes and the computation of auto-covariance matrices. Additionally, the lecture explores the recursive computation of auto-covariance matrices and the spectral representation of multivariate time series. The summary includes an overview of stationarity, linear processes, spectral decomposition, forecasting methods, financial time series models, and Kalman filtering.