We consider an unconstrained minimization problem with the objective function defined as a large sum of functions. Such problems arise in machine learning, stochastic optimization, data fitting and similar applications. In this paper we investigate inexact Newton method with subsampled Hessian. Numerical experiments on binary classification problems are presented
Inexact Newton methods for minimizing large sums / Stefania Bellavia, Natasa Krejic, Natasa Krklec Jerinkic. - ELETTRONICO. - (2017), pp. 0-0. (Intervento presentato al convegno 4th Conference on Optimization Methods and Software tenutosi a Cuba nel 16-20 Dicembre 2017).
Inexact Newton methods for minimizing large sums
Stefania Bellavia;
2017
Abstract
We consider an unconstrained minimization problem with the objective function defined as a large sum of functions. Such problems arise in machine learning, stochastic optimization, data fitting and similar applications. In this paper we investigate inexact Newton method with subsampled Hessian. Numerical experiments on binary classification problems are presented
File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
