Tutorials will take place on Friday the 18th of December 2020. Indications to access will be provided in due course.
This virtual tutorial will explain the theory and practice of forecasting when facing a nonstationary and evolving world, where the model differs from the data generation process (DGP). It covers the main sources of forecast error and explains how to produce forecasts following structural breaks, discussing how to robustify forecasts when there are shifts in distributions. Applications to empirical time series will demonstrate the approach. The OxMetrics software package will be used for the empirical applications.
Timetable, description of the sessions and bibliography can be found here.
Statistical methods play a crucial role in understanding and analyzing brain-imaging data. Bayesian approaches, in particular, have shown great promise in applications. A remarkable feature of Bayesian approaches is that they allow a flexible modeling of spatial and temporal correlations in the data. We will focus on the analysis and challenges provided by the use of fMRI data, although we will provide a quick review of some other common data types (e.g., EEG, PET/MRI, DTI). We will provide a general overview of the most relevant Bayesian modeling approaches developed in recent years, both from our group and others. We will divide methods according to the objective of the analysis. In particular, we will discuss spatio-temporal models for fMRI data that detect task-related activation patterns. We will also address the very important problem of estimating brain connectivity and touch upon methods that focus on making predictions of an individual's brain activity or a clinical or behavioral response. We also briefly discuss the emerging field of imaging genetics.