The Intelligent Systems Group at Paderborn University is seeking for
highly qualified doctoral or postdoctoral researchers interested in
machine learning. Candidates are expected to conduct research within a
project funded by the German Research Foundation. The contract is for
three years, and the payment is determined according to the competitive
German TVL E-13 scheme (depending on the candidate's experience and
qualifications). Within the project, there is a possibility for a
cooperation with Robert Busa-Fekete from Yahoo! Research, New York.
P O S I T I O N R E Q U I R E M E N T S
Ph.D. position applicants need to combine excellent skills in
mathematics, statistics, and computer science. A successful postdoc
applicant should have a strong background in machine learning with a
corresponding track record of research publications, including top-tier
conferences (e.g., ICML, NIPS, AISTATS, IJCAI, AAAI) and journals (e.g.,
JMLR, MLJ). Ideally, an applicant has experience on topics relevant for
the project (see below).
H O W T O A P P L Y
Ph.D. applicants should provide a research statement, their CV, degrees
including grade-sheets, and two references who are willing to write a
recommendation letter. Postdoc applicants should additionally provide
their top three publications. Please submit complete applications,
preferably combined in a single PDF file, to Prof. Eyke Hüllermeier
(eyke@upb.de). Please state the reference number 2847 in the subject.
There is no fixed deadline, but the positions will be filled as soon as
possible.
T H E P R O J E C T
In machine learning, the notion of multi-armed bandit (MAB) refers to a
class of online learning problems, in which an agent is supposed to
simultaneously explore and exploit a given set of choice alternatives in
the course of a sequential decision process. Combining theoretical
challenge with practical usefulness, MABs have received considerable
attention in machine learning research in the recent past. This project
is devoted to a variant of standard MABs that is referred to as the
dueling bandit or preference-based multi-armed bandit (PB-MAP) problem.
Instead of learning from stochastic feedback in the form of real-valued
rewards for the choice of single alternatives, a PB-MAB agent is allowed
to compare pairs of alternatives in a qualitative manner. The goal of
this project is to address several open research questions related to
the PB-MAB setting, and to study variants and extensions of this setting.
M O R E I N F O R M A T I O N
The homepage of the Intelligent Systems group can be found here:
https://www.cs.uni-paderborn.de/fachgebiete/intelligente-systeme.html