Second order optimization methods for the solution of large scale nonlinear noisy problems
Quand ? |
Le 14/10/2020, de 11:00 à 12:00 |
---|---|
Où ? | Salle de réunion du LIP, M7-316, ENS de Lyon, Site Monod |
Participants |
Elisa Riccietti |
Ajouter un événement au calendrier |
vCal iCal |
In this talk I will present new second order optimization methods to solve large scale problems with an expensive objective function, such as those arising in machine learning. We propose to exploit approximations of the objective function of dynamic accuracy to reduce the computational cost of the solution. We design two types of new second order methods (subsampled and multilevel) that are able to deal with the noise introduced by these approximations. We prove convergence and complexity guarantees for all methods and test them experimentally on real-life applications arising in particular in the classification of large data sets and deep neural network training.