Learning low-dimensional structure in Lagrangian systems

Prof. Daniel D. Lee (The University of Pennsylvania, Philadelphia, PA. , USA)

May 17, 2011 2:00 PM

302-309

 

Abstract:

 

Learning the nature of a physical system is a problem that presents many challenges and opportunities due to the unique structure associated with such systems. Many physical systems in engineering are high-dimensional, which prohibits the application of standard learning methods to such problems. We have recently worked on solving such learning problems by identifying their low-dimensional Lagrangian structure, and the associated inference problem of solving a high-dimensional minimum-cost path problem by exploiting the symmetry of the problem. These methods are demonstrated via application to learning structure in high-dimensional human motion and optimal motion planning for highly articulated robot mechanisms. s brain cell using the pattern matching techniques and also to develop an automation program which can detect contact points of two materials precisely in nano-scale.

 

 

This page is maintained by Jin-young Chun (jychun@bi.snu.ac.kr).
Last update: May 20, 2011