DYSCO Study Day – November 14, 2008
Palais des Académies / Paleis der Academiën
Rue Ducale, 1 / Hertogsstraat 1 1000 Brussels
9.30 Plenary lecture 1
Francis Bach, INRIA – Ecole Normale Supérieure, Paris
Multiple kernel learning refers to a theoretical and algorithmic framework aimed at learning the kernel matrix directly from data for supervised learning techniques such as the support vector machine. The framework is based on a convex parameterization of the set of kernel matrices and a convex formulation which can be cast as a block L1-norm regularization. In this talk, I will explore some applications and large-scale optimization algorithms, as well as some new links with sparsity-inducing norm theory.
10.30 to 11.45 Poster session 1
11.45 to 12.30 Status report of the first 3 Workpackages (WP)
12.30 to 14 Lunch
14 to 14.30 Status report of WP 4 and 5
14.30 to 15.45 Poster session 2
15.45 to 16.45 Plenary lecture 2
Georges Bastin, UCL
Using hyperbolic systems of balance laws for modeling, control and stability analysis of physical networks
Abstract: The operation of many physical networks having an engineering relevance may be represented by hyperbolic systems of balance laws in one space dimension. By using a Lyapunov stability analysis, it can be shown that the exponential stability of the equilibrium is guaranteed if the boundary conditions are dissipative. This essential property is helpful for solving the associated control problem of designing control laws at the network junctions in order to stabilize the system. This modelling and control design issue is illustrated with applications to flow control in hydraulic networks and congestion control in road traffic networks.