View source for Rodolphe Sepulchre, June 2013

Rodolphe Sepulchre will visit Caltech on 3 June 2013 (Mon).  

=== Agenda ===
{{agenda begin}}
{{agenda item|9:30a|Richard Murray, 109 Steele Lab}}
{{agenda item|9:45a|Meet with Richard's NCS group, 110 Steele}}
* 9:45-10:45: Necmiye, Mumu, Eric, Rangoli
* 10:45-11:45: Enoch, Marcella, Anandh, Dan
{{agenda item|11:45a|Lunch with Venkat, CMS faculty}}
{{agenda item|1:15p|Seminar setup}}
{{agenda item|1:30p|Seminar, 121 ANB}}
{{agenda item|3:00p|Venkat Chandrasekaran, 300 ANB}}
{{agenda item|3:45p|Lijun Chen, 202 ANB}}
{{agenda item|4:30p|(free), 320 ANB}}
{{agenda item|5:15p|Done}}
{{agenda end}}

=== Abstract ===

<center>
'''The geometry of (thin) SVD revisited for large-scale computations'''

Rodolphe Sepulchre<br>
University of Liege, Belgium
</center>

The talk will introduce a riemannian framework for large-scale  computations over the
set of low-rank  matrices. The foundation is geometric  and the motivation
is algorithmic, with a bias towards efficient computations in large-scale problems.
We will explore how classical matrix factorizations connect the riemannian geometry of the set of
fixed-rank matrices  to two well-studied manifolds: the Grassmann manifold of linear subspaces and the cone
of positive definite matrices. The theory will be illustrated on various applications, including 
low-rank Kalman filtering,  linear regression with low-rank priors, matrix completion,  and  the choice of a suitable metric for Diffusion Tensor Imaging.