A CLASS OF NONMONOTONE TRUST REGION ALGORITHMS FOR UNCONSTRAINED OPTIMIZATION PROBLEMS

Authors
KE XW
Citation
Xw. Ke, A CLASS OF NONMONOTONE TRUST REGION ALGORITHMS FOR UNCONSTRAINED OPTIMIZATION PROBLEMS, Science in China. Series A, Mathematics, Physics, Astronomy & Technological Sciences, 41(9), 1998, pp. 927-932
Citations number
11
Categorie Soggetti
Multidisciplinary Sciences
Journal title
Science in China. Series A, Mathematics, Physics, Astronomy & Technological Sciences → ACNP
ISSN journal
10016511
Volume
41
Issue
9
Year of publication
1998
Pages
927 - 932
Database
ISI
SICI code
1001-6511(1998)41:9<927:ACONTR>2.0.ZU;2-D
Abstract
A class of nonmonotone trust region algorithms is presented for uncons trained optimizations. Under suitable conditions, the global and Q-qua dratic convergences of the algorithm are proved. Several rules of choo sing trial steps and trust region radii are also discussed.