MPI-INF Logo
Campus Event Calendar

Event Entry

What and Who

Black-Box Optimization under Uncertainty

Dr. Timo Kötzing
Hasso-Plattner-Institut
AG1 Mittagsseminar (others' work)
AG 1, AG 2, AG 3, AG 4, AG 5, SWS, RG1, MMCI  
AG Audience
English

Date, Time and Location

Friday, 10 June 2016
14:00
60 Minutes
E1 4
024
Saarbrücken

Abstract

Black-box optimization considers optimization of problems which are only known via access to an objective function measuring the quality a given search point. Typical generic solvers to such problems include local search, but also many nature-inspired search heuristics such as Simulated Annealing and Evolutionary Algorithms. There is a growing community of theoretical computer scientists working on explaining the principles which make these search heuristics successful. In particular, recent progress shows that many of the considered approaches perform very well for optimization under uncertainty. In this talk I will discuss uncertainty stemming from noisy objective functions, where certain algorithms are particularly robust to noise, while others fail. I will show what distinguishes the successful algorithms from the less successful and present the used techniques (which deal with the analysis of stochastic processes) in a manner accessible to a wider audience

Contact

Christina Fries
--email hidden
passcode not visible
logged in users only

Christina Fries, 05/19/2016 10:00 -- Created document.