Francis Bach
Convex sparse methods for feature hierarchies.
Sparse methods usually deal with the selection of a few elements from
a large collection of pre-computed features. While theoretical results
suggest that techniques based on the L1-norm can deal with
exponentially many irrelevant features, current algorithms cannot
handle more than millions of variables. In this talk, I will show how
structured norms can deal in polynomial time with exponentially many
features that are organized in a directed acyclic graph.
Brief Bio.
Francis Bach is a researcher in the Willow INRIA project-team, in the Computer Science Department of the Ecole Normale Superieure, Paris, France. He graduated from the Ecole Polytechnique, Palaiseau, France, in 1997, and earned his PhD in 2005 from the Computer Science division at the University of California, Berkeley. His research interests include machine learning, statistics, optimization, graphical models, kernel methods, and statistical signal processing.