University of Technology, Sydney

Staff directory | Webmail | Maps | Newsroom | What's on
#

QCIS Seminar: A unifying learning framework in vector-valued reproducing kernel Hilbert spaces for vector-valued manifold regularisation, multi-view learning, and multi-task learning - Dr Hà Quang Minh, Istituto Italiano di Tecnologia (IIT), Italy

Presenter: Dr Hà Quang Minh, Istituto Italiano di Tecnologia (IIT), Genova, Italy (invited by Prof Dacheng Tao)

Abstract: Reproducing kernel Hilbert spaces (RKHS) hae recently emerged as a powerful mathematical framework for many problems in machine learning, statistics, and their applications. This talk will present a formulation in vector-valued RKHS that provides a unified treatment of several important machine learning approaches. The first is Manifold Regularisation, which seeks to exploit the geometry of the input data via unlabeled examples. The second is Multi-view Learning, which attempts to integrate different features and modalities in the input data. The third is Multi-task Learning, which attempts to learn many related tasks simultaneously. Numerical results on several challenging multi-class classification problems demonstrate the competitive practical performance of our approach.

Short bio: Hà Quang Minh received his BSc degree from Monash University and MSc and PhD degrees in mathematics (with Steve Smale) from Brown University, Providence, RI, USA, in May 2006. He is currently a Senior Post-Doctoral Researcher in pattern analysis and computer vision at the Istituto Italiano di Tecnologia (IIT), Genova, Italy. Prior to joining IIT, he worked at the University of Chicago, the University of Vienna (Austria), and  Humboldt University of Berlin (Germany). His current research interests include applied and computational functional analysis, machine learning, and applications in data analysis, image and signal processing.

 

 

Date:
2 December 2013
Time:
14:30 - 16:00
Location:
City - Broadway CB10 Level 2, Room 470
Audience:
All Welcome
Contact:
Barbara Munday

Back to list of Past Events