TITLE: Machines, Brains, Humans — and Other Computational Enigmas

ABSTRACT:

Computation — well-defined sequences of state changes — is universal. In this talk, we discuss recent struggles with understanding three aspects of it: (1) How does the brain create, recall, and associate memories? (Will you remember this abstract?) (2) How to measure the complexity of human computation? (Think adding in your head or playing chess) (3) What concepts/functions can (and cannot) be provably learned by deep neural networks? (This one might be worth some $$. )They all have a common answer — I don’t know. I’ll describe some surprises, several hypotheses, and a sea of challenges.

This abstract was generated by a recurrent NN; please excuse output errors.

BIO:

Santosh Vempala is a Distinguished Professor of Computer Science at the Georgia Institute of Technology. Vempala attended Carnegie Mellon University, where he received his Ph.D. in 1997 under Professor Avrim Blum. In 1997, he was awarded a Miller Fellowship at Berkeley. Subsequently, he was a professor at MIT in the Mathematics Department until he moved to Georgia Tech in 2006. His main work has been in the area of theoretical computer science, with particular activity in the fields of algorithms; randomized algorithms; computational geometry; and computational learning theory, including the authorship of books on random projection and spectral methods. In 2008, he co-founded the Computing for Good (C4G) program at Georgia Tech. Vempala has received numerous awards, including a Guggenheim Fellowship and Sloan Fellowship. He was named Fellow of ACM “for contributions to algorithms for convex sets and probability distributions” in 2015.

Watch the talk.