THESIS Jaquier_THESIS_2020/IDIAP Robot skills learning with Riemannian manifolds : Leveraging geometry-awareness in robot learning, optimization and control Jaquier, N. Bayesian optimization domain knowledge Learning from demonstrations manipulability ellipsoids Riemannian manifolds robot learning tensor methods EXTERNAL https://publications.idiap.ch/attachments/papers/2020/Jaquier_THESIS_2020.pdf PUBLIC 2020 Ecole Polytechnique Fédérale de Lausanne