Original title: A Generalized Limited-Memory BNS Method Based on the Block BFGS Update
Authors: Vlček, Jan ; Lukšan, Ladislav
Document type: Papers
Conference/Event: Programs and Algorithms of Numerical Mathematics /18./, Janov nad Nisou (CZ), 20160619
Year: 2017
Language: eng
Abstract: A block version of the BFGS variable metric update formula is investigated. It satisfies the quasi-Newton conditions with all used difference vectors and gives the best improvement of convergence in some sense for quadratic objective functions, but it does not guarantee that the direction vectors are descent for general functions. To overcome this difficulty and utilize the advantageous properties of the block BFGS update, a block version of the limited-memory BNS method for large scale unconstrained optimization is proposed. The algorithm is globally convergent for convex sufficiently smooth functions and our numerical experiments indicate its efficiency.
Keywords: block variable metric methods; global convergence; limited-memory methods; numerical results; the BFGS update; unconstrained minimization
Project no.: GA13-06684S (CEP)
Funding provider: GA ČR
Host item entry: Programs and algorithms of numerical mathematics 18, ISBN 978-80-85823-67-7
Note: Související webová stránka: http://dml.cz/handle/10338.dmlcz/703010

Institution: Institute of Computer Science AS ČR (web)
Document availability information: Fulltext is available in the digital repository of the Academy of Sciences.
Original record: http://hdl.handle.net/11104/0272937

Permalink: http://www.nusl.cz/ntk/nusl-358347

The record appears in these collections:
Research > Institutes ASCR > Institute of Computer Science
Conference materials > Papers
 Record created 2017-08-02, last modified 2021-11-24

No fulltext
  • Export as DC, NUŠL, RIS
  • Share