Nicole Megow: Teaching
Seminar Optimization under Uncertainty: Stochastic and Online Models (SS 2009)
Nicole Megow and
Rob van Stee
|
Time:
|
Tuesday 10-12
|
|
Room:
|
SR 016, building E13
|
|
Start:
|
Tuesday, April 21, 2009
|
|
Content:
|
In combinatorial optimization one typically assumes that the entire information about a problem instance is known in advance. However, when dealing with real-world optimization problems this is most often not the case. Think of the control of an elevator where the set of transportation requests is not known in advance or scheduling tasks in an operating system where the excecution time for the tasks are uncertain. There are two major frameworks modeling incomplete information in the theory of optimization: one deals with stochastic information, the other one with online information.
In this seminar we study recent research papers on interesting problems and fundamental techniques on how to cope with incomplete information when solving NP-hard problems.
|
|
Credit:
|
For a successful talk (60 min) and a written report (15-20 pages) you get 7 credit points (LP).
|
|
Registration:
|
If you are interested in this seminar, please send us an Email by Sunday, April 19.
|
|
Prerequisites:
|
Basic knowledge of algorithms and combinatorial optimization.
|
|
Schedule:
|
|
Date
|
Lecturer
|
Paper
|
|
May 26
|
Stephanie Ehrbächer
|
Boosted sampling: approximation algorithms for stochastic optimization
(A. Gupta, M. Pal, R. Ravi, A. Sinha)
|
|
June 2
|
Daniel Bach
|
Cancelled.
|
|
June 16
|
Faraz Makari
|
The power of reordering for online minimum makespan scheduling
|
|
June 23
|
Daniel Bach
|
Online dial-a-ride problems: minimizing the total completion time (N. Ascheuer, S.O. Krumke, J. Rambau )
|
|
July 14
|
Dag Sonntag
|
The online TSP against fair adversaries (M. Blom, S.O. Krumke, W. de Paepe, L. Stougie)
|
|