PhD positions at KTH Royal Institute of Technology (Stockholm, Sweden)

There are several PhD positions available in Artificial Intelligence at KTH Royal Institute of Technology in Stockholm.

Deadline: September 30th 2018

Detailed information:

https://www.kth.se/en/om/work-at-kth/lediga-jobb/what:job/jobID:219360/type:job/where:4/apply:1

Further information regarding project B):

Project B) may in particular be of interest also to students interested in Computational Topology/Geometry and Machine Learning as there are intriguing connections between e.g. computational geometry and recent machine learning approaches such as deep convolutional neural networks.

Data-Driven Foundations for Robust Deformable Object Manipulation

The project's goals are to develop new paradigms for machine learning under constraints imposed by physics. This is a key challenge with current approaches such as "Deep Learning" which often perform well empirically, but have to date few mathematical guarantees.

The student will in particular investigate how to combine approaches to rigorous mathematical modeling of deformable objects such as plastics and metal sheets with recent advances in machine learning for state estimation and prediction. The project will be conducted in collaboration also with Chalmers University where real-world experiments with industrial robots will be performed which will provide a possibility for validation of hypotheses and to test the applicability of the developed methods to controlling a robot's actions in deformable object manipulation tasks.

New ideas coming from geometry and topology, persistent homology and sheaf cohomology, etc may find application in developing new algorithms in this area and there is interest in candidates with a wide variety of backgrounds, including candidates with strong knowledge of computational geometry, algebraic geometry, topology, differential geometry, probability theory, etc.

For questions on project B), please contact:

Florian Pokorny,

Assistant Professor, KTH

http://www.csc.kth.se/~fpokorny

Leave a Reply

Your email address will not be published. Required fields are marked *

 

This site uses Akismet to reduce spam. Learn how your comment data is processed.