Al fine di migliorare la tua esperienza di navigazione, questo sito utilizza i cookie di profilazione di terze parti. Chiudendo questo banner o accedendo ad un qualunque elemento sottostante acconsenti all’uso dei cookie.

Asynchronous parallel algorithms for the minimization of nonsmooth, nonconvex functions.

30 novembre 2017 ore 12:01 - 13:00

Aula 207, Sede di Viale Romania, 32

Speaker: Francisco Facchinei, Sapienza, University of Rome

We present  recent advancements in the field of asynchronous parallel methods for the minimization of the sum of a differentiable function and a possibly nonsmooth, convex regularizer subject to constraints. In recent years, problems of this kind have played an important role in many applicative fields, e.g. in machine learning,  data mining, and compressed sensing, and instances of ever increasing dimensions need to be solved. Asynchronous methods play a key role in the solution of these large problems. After reviewing recent results in the field, we discuss, in particular, a rather general framework  along with its main convergence properties  and  report numerical results.