<?xml version="1.0" encoding="ISO-8859-1" ?>
<rss version="2.0"
     xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"  
     xmlns:ev="http://purl.org/rss/1.0/modules/event/" 
     xmlns:dc="http://purl.org/dc/elements/1.1/"
>
 <channel>
  <title>URCS Seminars &amp; Talks</title>
  <link>http://www.cs.rochester.edu/dept/seminar</link>
  <description>Upcoming Seminars &amp; Talks from the Computer Science Department at the University of Rochester</description>
  <language>en-us</language>
  <image>
   <url></url>
   <title>Computer Science Department, University of Rochester</title>
   <link>http://www.cs.rochester.edu/dept/seminar</link>
  </image>
  <lastBuildDate>Wed, 17 Mar 2010 09:19:55 -0400</lastBuildDate>
  
   <item>
    <title>Vyas Sekar: Effective Management through System-Wide Coordination and Optimization</title>
    <link>http://www.cs.rochester.edu/dept/seminar/view/488</link>
    <description>[Monday, March 22, 2010 at 11:00 AM in Computer Studies Bldg. Room 209] As networked systems grow and traffic patterns evolve, management applications
are increasing in complexity and functionality. To address these demands,
equipment vendors and administrators today depend on incremental solutions that
increase the complexity of network elements and deployment costs for operators.
Despite this increased complexity and cost, there is still a fundamental
disconnect between the policy objectives of system administrators and today&#39;s
mechanisms.

I argue that much of this disconnect arises from the narrow device-centric view
of current solutions. Such piecemeal solutions are inefficient: network
elements duplicate tasks and some locations become overloaded.  Worse still,
administrators struggle to retrofit their high-level goals within
device-centric configurations.  My dissertation argues for a clean-slate
system-wide approach for resource management in large-scale networked systems
based on three high-level principles: (1) systematic selection and placement of
device-level primitives, (2) lightweight coordination mechanisms that enable
different network elements to operate in synchrony, and (3) practical
optimization models that capture operating constraints and policy objectives.
In this talk, I will highlight the benefits of such a system-wide approach in
the context of meeting fine-grained coverage and accuracy requirements in
traffic monitoring and implementing a redundancy elimination service to improve
network performance. Further, these architectural principles can be more
broadly applied in other contexts such as intrusion detection and prevention,
multi-hop wireless and sensor networks, managing cloud services, and performance
troubleshooting in data centers. 



Bio: Vyas Sekar is a final year PhD student in the Computer Science Department at
Carnegie Mellon University, co-advised by Michael Reiter and Hui Zhang.  His
research interests are in systems, networking, and security. Before CMU, he
earned his bachelor&#39;s degree from the Indian Institute of Technology Madras,
where he was awarded the President of India Gold Medal.
 

Refreshments will be provided at 10:45AM
</description>
    <ev:startdate>20100322T150000Z</ev:startdate>
    <ev:enddate>20100322T160000Z</ev:enddate>
    <guid>http://www.cs.rochester.edu/dept/seminar/view/488</guid>
   </item>
  
   <item>
    <title>Bryan Russell: LabelMe: online image annotation and applications</title>
    <link>http://www.cs.rochester.edu/dept/seminar/view/486</link>
    <description>[Friday, March 26, 2010 at 10:00 AM in Computer Science Bldg. Room 601] Central to the development of computer vision systems is the
collection and use of annotated images spanning our visual world.
Annotations may include information about the identity, spatial
extent, and viewpoint of the objects present in a depicted scene. Such
a database is useful for the training and evaluation of computer
vision systems. Motivated by the availability of images on the
internet, we introduced a web-based annotation tool that allows online
users to label objects and their spatial extent in images. To date, we
have collected over 500K annotations that span a variety of different
scene and object classes.  In this talk, I will show the contents of
the database, its growth over time, and statistics of its usage. In
addition, we use the collected user-provided object annotations to
extract the real-world 3D coordinates of images in a variety of
scenes. Important for this task is the recovery of geometric
information that is implicit in the object labels, such as qualitative
relationships between objects (attachment, support, occlusion) and
quantitative ones (inferring camera parameters).  We show that we are
able to obtain high quality 3D information by evaluating the proposed
approach on a database obtained with a laser range scanner.

Bio: Bryan Russell is a postdoctoral fellow in the INRIA Willow team at the
Ecole Normale Superieure in Paris, France.  His research is in the
area of computer vision, with particular interests in object
recognition and scene understanding.  Bryan received his PhD in 2008
from MIT.

Refreshments provided at 9:45AM
</description>
    <ev:startdate>20100326T140000Z</ev:startdate>
    <ev:enddate>20100326T150000Z</ev:enddate>
    <guid>http://www.cs.rochester.edu/dept/seminar/view/486</guid>
   </item>
  
   <item>
    <title>Zoya Svitkina: Asymmetric Traveling Salesman Path and Directed Latency Problems</title>
    <link>http://www.cs.rochester.edu/dept/seminar/view/489</link>
    <description>[Monday, March 29, 2010 at 11:00 AM in Computer Studies Bldg. Room 209] The Traveling Salesman Problem (TSP) is one of the best-known problems
in combinatorial optimization, with numerous applications. In this
talk I present recent work that studies integrality gaps and
approximability of two generalizations of TSP on directed graphs.
Given two specified nodes s and t, both problems ask to find an s-t
path visiting all other nodes. In the asymmetric traveling salesman
path problem (ATSPP), the objective is to minimize the total cost of
this path. In the directed latency problem, the objective is to
minimize the sum of distances on this path from s to each node. This
can model, for example, minimizing the average waiting time of
customers located at the nodes. I present a new algorithm and
integrality gap bound for ATSPP and their use for designing an
approximation algorithm for the directed latency problem.

Bio:  Zoya Svitkina has obtained a Bachelor of Science degree from the
University of Wisconsin - Madison, and a Ph.D. in Computer Science
from Cornell University, under the supervision of Prof. Eva Tardos.
She is currently a Postdoctoral Fellow at the University of Alberta in
Canada.  Her research interests are in theoretical computer science,
combinatorial optimization, and approximation algorithms.

Refreshments will be provided at 10:45 AM
</description>
    <ev:startdate>20100329T150000Z</ev:startdate>
    <ev:enddate>20100329T160000Z</ev:enddate>
    <guid>http://www.cs.rochester.edu/dept/seminar/view/489</guid>
   </item>
  
   <item>
    <title>Ian Kash: Algorithmic Market Design: Currency Crashes, Spectrum Sales, and Lying Hospitals</title>
    <link>http://www.cs.rochester.edu/dept/seminar/view/490</link>
    <description>[Monday, April 05, 2010 at 11:00 AM in Computer Studies Bldg. Room 209] Markets are a powerful tool for producing desirable outcomes in an
engineered system.  As computers allow us to build new and more
complex markets, we need new theory and algorithms to understand,
build, and improve them.  In this talk I&#39;ll cover three examples of
such markets, involving fake money, real money, and no money at all.

Scrip systems, where users pay for service with an artificial currency
(scrip) created for the system, are an attractive solution to free
riding and other problems in P2P and distributed systems. Despite the
interest in building scrip systems, relatively little work has been
done to help answer basic design questions. For example, how much
money should there be in the system? What will happen if some of the
users start giving away service or hoarding money?  I&#39;ll show how a
simple algorithm provides insight into these questions.

Until recently, if you wanted to start a business that requires using
wireless spectrum, you would either have to pay tens of millions of
dollars for an exclusive license or use one of the few congested
frequencies that are available without a license (such as the one used
by WiFi).  Recently, people have begun investigating how auctions can
be used to sell smaller amounts of spectrum.  I&#39;ll present a new
auction algorithm that takes advantage of the ability of devices to
share spectrum to significantly improve both spectrum utilization and
social welfare.

Eventually, patients with kidney disease need a new kidney.  Luckily,
we can live with only one kidney so donation is possible, so many
patients are able to receive a new kidney from a family member or
friend.  However, not all donors and patients are compatible.  A
recent innovation is the idea of a kidney exchange, where two donors
who are incompatible with their intended patients instead supply a
kidney to the other&#39;s patient.  Hospitals have begun to participate
in programs to organize exchanges, but not all of them truthfully
report all of their available patients.  I&#39;ll introduce a new matching
algorithm that removes the incentive for hospitals to lie.

Bio: Ian Kash is a postdoctoral fellow at the Center for Research on
Computation and Society at Harvard University.  He received his
B.S. from Carnegie Mellon University and his Ph.D. from Cornell
University.  His thesis, &quot;Robustness and Optimization of Scrip
Systems,&quot; was advised by Eric Friedman and Joseph Halpern.  His
research interests include market design, (algorithmic) game theory,
theoretical aspects of systems, and multi-agent learning.

Refreshments will be provided at 10:45 AM</description>
    <ev:startdate>20100405T150000Z</ev:startdate>
    <ev:enddate>20100405T160000Z</ev:enddate>
    <guid>http://www.cs.rochester.edu/dept/seminar/view/490</guid>
   </item>
  
 </channel>
</rss>
