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 by the instructor covers the continuation of prediction in time series, focusing on ARMA models. It explains the forecasting process for general ARMA processes, the implications of the representation, and the properties of the forecast error. The lecture also delves into forecasting an ARMA(1,1) model, discussing the model equation, calculating forecasts, and determining the forecast error. Additionally, it explores the challenges with predictions, error measurements, and the importance of selecting the right error metric. The lecture concludes with a discussion on covariance models, long-memory processes, and fractionally differenced processes.