[Comp-neuro] 2 PhD Positions in Computational Neuroscience

Daniel Braun daniel.braun at tuebingen.mpg.de
Tue Jul 12 16:49:19 CEST 2011


2 PhD Positions in Computational Neuroscience

Max Planck Institute for Biological Cybernetics

There are two fully funded PhD positions available in the newly
established Emmy-Noether group on “Computational and Biological
Principles for Sensorimotor Learning” at the Max-Planck-Institute for
Biological Cybernetics. The group is interested in the mathematical
foundations of intelligent behaviour in biological and artificial
systems. Following the notion of embodied intelligence, we test
mathematical hypotheses experimentally in human sensorimotor control
using virtual reality technology. The group is also interested in
applications that allow sensorimotor learning in robotic devices.
Thus, we have a combined theoretical and experimental approach. The
two PhD positions will be thematically centred on two cutting-edge
problems:

a) The problem of structural learning / abstraction

When humans (and other animals) learn to solve problems, they
typically learn much more than just solving one particular task. This
is evidenced when they use their experience to quickly solve novel
problems that are somehow “related”. A powerful mechanism of such
generalization is the extraction of structural invariants that govern
a whole range of tasks with certain structural features. The ability
to abstract structural similarities and causal interdependencies in
variable environments is a hallmark of intelligent behaviour and one
of the hardest problems in artificial intelligence research.
Extracting invariants in the sensorimotor stream provides a mechanism
for “learning-to-learn” and might also ground higher forms of concept
formation. The aim of this PhD project is to gain a quantitative
understanding of structural learning processes in sensorimotor
behaviour.

b) The problem of bounded rationality (neuroeconomics)

Intelligent (or rational) behaviour is often described mathematically
by optimality principles both in economics and in biology. However,
all organisms have finite (bounded) computational resources and have
to react quickly to changes in their environment without wasting too
much time on finding the absolute optimal strategy. Similarly,
intelligent machines need to consider their computational limitations
when interacting with the world in real time. Models of bounded
rationality take such resource limitations into account and constitute
one of the most important research topics at the forefront both in
artificial intelligence research and behavioural economics. The aim of
this project is to develop models of bounded rational sensorimotor
control and to study in a quantitative manner how bounded rationality
affects sensorimotor learning and behaviour.

The PhD positions are paid according to German tariff TVöD E13 50%
(ca. €1100,- per month after tax for a single person). The
Emmy-Noether group is funded by an excellence programme of the
Deutsche Forschungsgemeinschaft (DFG). The setting in the
Max-Planck-Institute provides a unique research environment in the
picturesque medieval town of Tübingen, Germany. There will be plenty
of opportunity to cooperate with laboratories in robotics,
neurophysiology, imaging, computational neuroscience and human
psychophysics. There will also be opportunities to work with
international partners in the UK and USA. The applicants should have a
strong mathematical background, hold a degree in natural science,
mathematics, economics, psychology, computer science, engineering or
similar and have a keen interest in neuroscience and artificial
intelligence.

The Max Planck Society is an equal opportunity employer: Handicapped
individuals are strongly encouraged to apply, and so are women in
areas in which they are underrepresented.

Candidates should send a CV and a brief statement of interest to Dr
Daniel A. Braun daniel.braun at tuebingen.mpg.de


More information about the Comp-neuro mailing list