Skip to main navigation Skip to Content

Computer Science

University of Toronto
  • U of T Portal
  • Site Map
  • Contact
  • About DCS At U of T
    • Why Study CS at U of T
    • Career Options
    • History of DCS
    • Giving to DCS
    • Information for Prospective Undergraduate Students
    • Information for Prospective Graduate Students
    • Computer Science at UofT Mississauga
    • Computer Science at UofT Scarborough
    • Contact
  • Programs & Courses
    • Prospective Undergraduate Students
    • Current Undergraduate Students
    • Prospective Graduate Students
    • Current Graduate Students
  • Research
    • Research Groups
    • Industrial Relations
    • Research In Action Showcase
    • Research Profiles
    • Research Sponsors & Partners
    • Awards and Accolades
    • UTRECS - Undergraduate Toronto Research Experience in Computer Science
  • Our People
    • Faculty
    • Staff
    • In Memoriam
    • People Profiles
    • Alumni and Friends
    • Women in Computer Science
    • Graduate Student Society
    • Undergraduate Student Union
    • Undergraduate Artificial Intelligence Group
  • News & Events
    • Current News
    • DCS Events Calendar
    • DCS in the Media
    • @dcs Newsletter
    • Undergrad News
    • Distinguished Lecture Series
    • Awards and Accolades
    • DCS Facebook Page
    • DCS Twitter Feed
    • RSS Feed - News
    • RSS Feed - Events
You are viewing: > Home > News & Events > DCS Events Calendar > Theory Seminar - June 24
  • Current News
  • DCS Events Calendar
  • DCS in the Media
  • @dcs Newsletter
  • Undergrad News
  • Distinguished Lecture Series
  • Awards and Accolades
  • DCS Facebook Page
  • DCS Twitter Feed
  • RSS Feed - News
  • RSS Feed - Events

Theory Seminar - June 24

Event date: Friday, June 24, 2011, at 11:10 AM
Location: GB 244

SPEAKER: Pascal Koiran
Ecole Normale Superieure de Lyon
University of Toronto

TITLE: Random graphs without large dense subgraphs: are they hard to certify?

ABSTRACT:
Given a graph on n vertices, is there a subgraph induced by k vertices with edge density at least 0.5+epsilon ?
For suitable values of the parameters k and epsilon, the answer is NO for most graphs. It then makes sense to ask for an algorithm certifying that most graphs do not contain any dense k-subgraph. We propose efficient algorithms for certain values of the parameters, and identify other values for which this problem seems hard. This work is motivated by applications to compressed sensing, and more precisely to the computational complexity of the restricted isometry
property (no background in this area will be assumed).

This is joint work with Anastasios Zouzias.

Computer Science

All rights reserved copyright Computer Science, University of Toronto