by Marco Congedo, CNRS, University Grenoble Alpes

Abstract: The Riemannian geometry of positive definite matrices has
allowed the introduction of state-of-the-art classification methods for
brain-computer interface (BCI) data. The use of this framework is
steadily increasing in the BCI community, sustained by its excellent
classification accuracy, robustness and ability to operate transfer
learning. In this talk the fundamental tools for manipulating and
classifying BCI data in the Riemannian manifold of positive definite
matrices will be introduced and explained. Particular emphasis will be
given to practical aspects of their use, referencing to available free
code resources in the Julia, Python, Matlab, R and C++ programming
languages.