[robotics-worldwide] [jobs] Postdoc position in the field of interactive machine learning at CoR-Lab, Bielefeld University

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

[robotics-worldwide] [jobs] Postdoc position in the field of interactive machine learning at CoR-Lab, Bielefeld University

For the newly started project cluster CINEMENTAS - Cooperative
Interaction based on Mental Models for Assistive Systems we solicit
several research positions with salary paid according to german TVL-13.
Project duration is 36 months.

CINEMENTAS is focused on the research question how learning by a
technical system in interaction with a human can be shaped to remain
transparent for the human user. To achieve this goal CINEMENTAS connects
machine learning approaches with concepts from cognitive science, such
as mental models.

CINEMENTAS is located at the Institute for Cognition and Robotics
(CoR-Lab) and embedded in the research environment of the Cluster of
Excellence Cognitive Interaction Technology (CITEC), both at Bielefeld
University, Germany. Project work will be in close cooperation with the
Honda Research Institute Europe (HRI, located in Offenbach near
Frankfurt/Main), which also provides the financial funding of the project.

The project is consists of five subprojects affiliated with five
research groups at the Faculty of Technology and at the Faculty of
Linguistics and Literary Science at Bielefeld University. The present
PostDoc position is dedicated to CINEMENTAS subproject Cognitive
Corridor Learning (CoCoLea, Prof. Dr. Helge Ritter, AG Neuroinformatics).

Aim of CoCoLea is the development of innovative strategies and
algorithms that allow to  constrain interactive machine learning within
a "corridor of readability" for the human user. Underlying idea is to
exploit the redundancy of the solution space for a learning task and to
benefit from cognitive science concepts, in particular mental models, to
shape learning towards maximal understandability for the human.

Scope of functions:

Research activities (85 %):

  - Development and implementation of algorithms to realize interactive
machine learning within a corridor of readability for the human user.
  - Technical realization of a prototype system within the chosen
demonstrator scenario and environment
  - Design and conduct of empirical user studies for data acquisition
and evaluation of the model/system

Other acticvities (15 %):

  - Reporting and collaboration with project partners, especially HRI
  - Preparation of scientific publications and presentations
  - If necessary and thematically related to the project's research
topic, supervision of Bachelor/Master theses and student projects

Applicants are required to have a qualified PhD in a pertinent topic
area within the fields of machine learning or human-machine interaction,
and an excellent diploma or master degree in informatics/computer
science or in a pertinent neighbor discipline, obtained at a research

We expect

  - solid practical and theoretical expertise in machine learning,
particularly in the context of interactive settings
  - a good publication record in pertinent international conferences
and/or scientific journals
  - excellent programming skills (C++,Python,Matlab),
  - capability for interdisciplinary cooperation
  - team orientation and a strong motivation to integrate results in the
form of implemented software components in a shared demonstrator system
  - good command of English in writing and speaking

We further appreciate a background in cognitive science concepts about
learning and mental representations
and a keen interest and willingness to learn across discipline borders.

Work in the project will proceed in close connection with four other
CINEMENTAS projects (cf. connected advertisements)

Mental models for collaborative reinforcement learning (Prof. Dr. Stefan
Kopp, AG Cognitive Systems and social Interaction)
Mental models of the validity of adaptive systems (Prof. Dr. Barbara
Hammer, AG   Machine Learning)
Collaborating on learning to classify (Prof. Dr. David Schlangen, AG
Applied Computer Linguistics)
Learning of co-constructed interactive signals in joint tasks (Prof. Dr.
Britta Wrede,   AG Applied Computer Science)

Research work in the project will include temporary stays at the Honda
Research Institute Europe located in Offenbach near Frankfurt/Main
(about 4h train distance from Bielefeld)

Please send your application, preferably by  E-Mail in a single
pdf-document no later than
Oct-20 2017 to:

Prof. Dr. Helge Ritter
Secretary Mrs. Susanne Strunk [hidden email]
Faculty of Technology
Bielefeld University
Postfach 100131
33501 Bielefeld

Questions can be directed to Prof. Dr. Helge Ritter
[hidden email]

robotics-worldwide mailing list
[hidden email]