Connectionists: Postdoctoral Position at UC San Diego - Physically Inspired Machine Learning

[Todd Hylton]

Because real-world environments are too complex to be effectively
understood and captured by present day engineering models, current
technologies are largely incapable of interacting with the real-world
without substantial human assistance.  Preferably, technologies would
evolve solutions to high-level tasks that are easily communicated by
humans, but the concept of generic “machine evolution” is largely
unexplored.  In this post-doctoral position, the researcher will develop
physically-inspired machine learning algorithms for systems (e.g. robots)
interacting with a physical environment.  Such algorithms should, for
example, incorporate online, unsupervised learning of multiscale
spatio-temporal structure and emphasize “learning the physics of / physical
interaction with” real world environments.  The goals are developing a
unified conceptual understanding grounded in physics and demonstrating key
concepts in simulation.  Applicants should have a strong background in
statistical physics, thermodynamics and dynamical systems as well as
expertise in computing and machine learning.  The position reports to
Prof. Todd
Hylton <https://www.linkedin.com/in/todd-hylton-3841b81>, is located within
the Contextual Robotics Institute
<http://jacobsschool.ucsd.edu/contextualrobotics/> at UC San Diego, and
will interact with other researchers and faculty working in robotics,
machine learning and neuromorphic computing.  Please send your CV and a
short statement of interest to thylton at eng.ucsd.edu <thylton at ucsd.edu>.

-- 
Regards,

Todd Hylton <https://www.linkedin.com/in/todd-hylton-3841b81>
Executive Director, Contextual Robotics Institute
<http://jacobsschool.ucsd.edu/contextualrobotics/>
Professor of Practice, Department of Electrical and Computer Engineering
<http://ece.ucsd.edu/>
Jacobs School of Engineering <http://jacobsschool.ucsd.edu/>
UC San Diego
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20170302/38300be1/attachment.html>