Algorithms for Solving Optimization Problems Arsing from Deep Neural Net Models: Smooth Problems

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F16%3A00309135" target="_blank" >RIV/68407700:21230/16:00309135 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Algorithms for Solving Optimization Problems Arsing from Deep Neural Net Models: Smooth Problems

Popis výsledku v původním jazyce

Machine Learning models incorporating multiple layered learning networks have been seen to provide e_ective models for various classi_cation problems. The resulting optimization problem to solve for the optimal vector minimizing the empirical risk is, however, highly nonlinear. This presents a challenge to application and development of appropriate optimization algorithms for solving the problem. In this paper, we summarize the primary challenges involved and present the case for a Newton-based method incorporating directions of negative curvature, including promising numerical results on data arising from security anomally deetection.

Název v anglickém jazyce

Algorithms for Solving Optimization Problems Arsing from Deep Neural Net Models: Smooth Problems

Popis výsledku anglicky

Machine Learning models incorporating multiple layered learning networks have been seen to provide e_ective models for various classi_cation problems. The resulting optimization problem to solve for the optimal vector minimizing the empirical risk is, however, highly nonlinear. This presents a challenge to application and development of appropriate optimization algorithms for solving the problem. In this paper, we summarize the primary challenges involved and present the case for a Newton-based method incorporating directions of negative curvature, including promising numerical results on data arising from security anomally deetection.

Druh

V_souhrn - Souhrnná výzkumná zpráva

CEP obor

JC - Počítačový hardware a software

OECD FORD obor

—

Projekt

—

Návaznosti

N - Vyzkumna aktivita podporovana z neverejnych zdroju

Rok uplatnění

2016

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Počet stran výsledku

5

Místo vydání

Praha

Název nakladatele resp. objednatele

CISCO SYSTEMS (Czech Republic) s.r.o.

Verze

—