Free Probability and Operators
[an error occurred while processing this directive]
|
Date Time |
Location | Speaker |
Title – click for abstract |
 |
01/24 4:00pm |
BLOC 306 |
Carl Pearcy Texas A&M University |
Does the existence of nontrivial invariant subspaces imply that of nontrivial hyperinvariant subspaces for operators on Hilbert space?
In this lecture I will discuss the status of my conjecture that operators with n.i.s. have n.h.s. and various sufficient conditions that imply that the conjecture is true |
 |
02/14 4:00pm |
BLOC 306 |
Benjamin Eisley UC Berkeley |
Neural Hessians - An Application of Free Probability
The empirical spectral distribution (ESD) of the loss function of a neural network contains important qualitative information about the learning process and can be used to estimate the rate at which the loss approaches its minima. However, the ESD is typically very difficult to compute. Generalizing work of Liao and Mahoney, we have developed a recursive algorithm to estimate the ESD of a single layer network with many-dimensional output. In this talk, I will introduce the concepts underlying our research, and discuss avenues for future study.
This talk is based on joint work with Aren Martinian, under the guidance of Professor Federico Pasqualotto. The research was partially funded by the SURF summer fellowship for Berkeley students. |
 |
02/28 4:00pm |
BLOC 306 |
Carl Pearcy Texas A&M University |
(POSTPONED, new date TBA) Toward a solution of my conjecture that n.i.s implies n.h.s.
In this lecture I will continue my talk of last time presenting various sufficient conditions that the conjecture be true. |
[an error occurred while processing this directive]