[Colloquium] Theory Seminars at Computer Science

[Donna Brooms]

Combinatorics & Theoretical Computer Science Seminar

 
Shmuel Weinberger
University of Chicago
 
Tuesday, June 2, 2015
3:00 pm
Ryerson 251

Title: “Manifold learning, complexity, random fields and disordered solids”
 
Abstract: Manifold learning is devoted to finding geometric structure in high dimensional data (with a goal towards exploiting this structure in further analyses).  After a rapid overview of some of this area, I will show how there are some phase transitions in (sample, Kolmogorov and logical) complexity for a few toy problems due to dimension or singularities.  Two responses to this problem come via the theory of Gaussian random fields or via a variation of the theory of property testability in graph theory (that has already been partially developed by theoretical physicists e.g. for understanding quasicrystals and semiconductors at low temperatures).

Host: Prof. Razborov 
 
(Refreshments will be served prior to the talk in Ry. 255 @ 2:30 p.m. )
 
 
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20150527/53cf75b9/attachment.htm