BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Total Variation Data Analysis
DTSTART:20131119T110000
DTEND:20131119T120000
DTSTAMP:20260407T051038Z
UID:ffde55826d3b2aeabedefe36701b88c344ef92e8ce7f6b0c0c982a70
CATEGORIES:Conferences - Seminars
DESCRIPTION:Xavier Bresson\nBio: Xavier received a B.A. of Physics from Un
 iversity of Marseille\, a M.Sc. in Electrical Engineering from Ecole Super
 ieure d’Electricite (SUPELEC) in Paris and a M.Sc. in Signal Processing 
 from University of Paris XI. In 2005\, he completed a Ph.D. at the Swiss F
 ederal Institute of Technology (EPFL). In 2006-10\, he joined the Departme
 nt of Mathematics at University of California\, Los Angeles (UCLA) as a Po
 stdoctoral Scholar. In 2010-13\, he was with the Department of Computer Sc
 ience at City University of Hong Kong as an Assistant Professor. In 2013\,
  he joined the Center for Biomedical Imaging (CIBM) and the Medical Image 
 Analysis Laboratory (MIAL) of the University of Lausanne (UNIL).\nWith eno
 rmous and daily flows of data in finance\, security\, health\, social netw
 ork and multimedia (sound/text/image/video)\, there is a strong need to pr
 ocess information as efficiently as possible for smart decisions to be mad
 e. Machine Learning develops analytical methods and strong algorithms to d
 eal with this massive large-scale\, multi-dimensional and multi-modal data
 . This field has recently seen tremendous advances with the emergence of n
 ew powerful techniques combining the key mathematical tools of sparsity\, 
 convex optimization and relaxation methods. In this talk\, I will present 
 how these concepts can be applied to find excellent approximate solutions 
 of NP-hard balanced cut problems for unsupervised data clustering\, signif
 icantly overcoming state-of-the-art spectral clustering methods including 
 Shi-Malik's normalized cut. I will also show how to design fast algorithms
  for the proposed non-convex and non-differentiable optimization problems 
 based on recent breakthroughs in total variation optimization problems bor
 rowed from the compressed sensing field. This new total variation clusteri
 ng technique paves the way to a new generation of learning algorithms that
  can provides simultaneously accurate\, fast and robust solutions to other
  fundamental problems in data science such as Support Vector Machine data 
 classification. These new methodologies have a wide range of applications 
 including data retrieval (search engines)\, neuroimaging (diseases detecti
 on and analysis) and social network analysis (community detection).
LOCATION:INM11
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR