[robotics-worldwide] CfP: NIPS 2013 Workshop on Advances in Machine Learning for Sensorimotor Control

Tom Walsh twalsh at mit.edu
Tue Aug 20 12:57:07 PDT 2013


NIPS 2013 Workshop on Advances in Machine Learning for Sensorimotor Control

Call for Papers

When/Where: A one day (Dec. 9th or 10th) workshop at NIPS, Lake Tahoe,
Nevada, USA

Web: http://acl.mit.edu/amlsc/nips13-Workshop/Main.html

Description:

Various sensorimotor frameworks have been effective at controlling
physical and biological systems, but many techniques rely on
pre-specified models to derive useful policies. Advances in machine
learning, including non-parametric Bayesian modeling/inference and
reinforcement learning allow systems to learn better models and
policies from data. However, incorporating modern machine learning
techniques into sensorimotor control systems can be challenging due to
the learner's underlying assumptions, the need to model uncertainty,
and the scale of such problems. This workshop will bring together
researchers from machine learning, control, and neuroscience that
bridge this gap between effective planning systems and machine
learning techniques to produce better sensorimotor control. Domains of
interest include autonomous robots and vehicles, as well as complex
real world systems, such as neural control or healthcare where actions
may take place over a longer timescale.


Relevant Topics:

- Integrating machine learning and planning/control
- Scaling machine learning techniques for real physical and biological systems
- Dealing with uncertainty in planning and control
- Exploration/Exploitation tradeoffs
- Machine learning for high frequency data
- Porting successful supervised or unsupervised learning techniques to
sensorimotor control
- Leveraging expert knowledge, demonstrations or priors in learning and planning
- Safety and risk sensitivity in planning and learning
- Modeling, planning, and control under uncertainty in biological systems
- Transferring biological insights to mechanical systems
- Engineering insights with a biological explanation
- Shared lessons between the control, neuroscience, and reinforcement
learning communities

Submission Details:

Authors are encouraged to submit their related work to the workshop by
9th of October 11:59 PM PDT (UTC -7 hours) in NIPS format. Submissions
should be a maximum of 8 pages with an extra page for references,
though shorter papers are welcome as well.  Submissions do not need to
be anonymous.  Papers can be submitted through the EasyChair website
https://www.easychair.org/conferences/?conf=amlsc13


Important Dates:

Submission - 9 October 2013 11:59 PM PDT (UTC -7 hours)
Notification - 23 October 2013
Workshop – 9th or 10th December 2013


Organizers:

Thomas Walsh
Alborz Geramifard
Marc Deisenroth
Jonathan How
Jan Peters

If you have questions about the workshop or the submission process
please contact Thomas Walsh [twalsh AT mit.edu]


More information about the robotics-worldwide mailing list