Limited-Memory Variable Metric Methods that use Quantities from the Preceding Iteration

0358545 - ÚI 2012 RIV CZ eng C - Konferenční příspěvek (zahraniční konf.)
Vlček, Jan - Lukšan, Ladislav
Limited-Memory Variable Metric Methods that use Quantities from the Preceding Iteration.
Programs and Algorithms of Numerical Mathematics. Praha: MÚ AV ČR, v. v. i., 2010 - (Vejchodský, T.; Chleboun, J.; Přikryl, P.; Segeth, K.; Šístek, J.), s. 227-232. ISBN 978-80-85823-57-8.
[Programs and Algorithms of Numerical Mathematics /15./. Dolní Maxov (CZ), 06.06.2010-11.06.2010]
Grant CEP: GA ČR GA201/09/1957
Výzkumný záměr: CEZ:AV0Z10300504
Klíčová slova: optimization * variable matrix method * limited-memory method * matrix-free mehods
Kód oboru RIV: BA - Obecná matematika
http://www.math.cas.cz/~panm/Panm15/proceedings/PANM15proc.pdf

In this contribution, a new family of globally convergent limited-memory (LM) variable metric (VM) line search methods for unconstrained minimization is presented. Numerical results indicate that the new methods can save computational time substantially for certain problems in comparison with the well-known L-BFGS method.
Trvalý link: http://hdl.handle.net/11104/0196548