Iranian Journal of Mathematical Sciences and Informatics
مجله علوم ریاضی و انفورماتیک
IJMSI
Basic Sciences
http://ijmsi.ir
1
admin
1735-4463
2008-9473
8
10.61186/ijmsi
14
8888
13
en
jalali
1400
1
1
gregorian
2021
4
1
16
1
online
1
fulltext
en
A Trust-region Method using Extended Nonmonotone Technique for Unconstrained Optimization
عمومى
General
پژوهشي
Research paper
<div dir="ltr">In this paper, we present a nonmonotone trust-region algorithm for unconstrained optimization. We first introduce a variant of the nonmonotone strategy proposed by Ahookhosh and Amini cite{AhA 01} and incorporate it into the trust-region framework to construct a more efficient approach. Our new nonmonotone strategy combines the current function value with the maximum function values in some prior successful iterates. For iterates far away<br>
from the optimizer, we give a very strong nonmonotone strategy. In the vicinity of the optimizer, we have a weaker nonmonotone strategy. It leads to a medium nonmonotone strategy when iterates are not far away from or close to the optimizer. Theoretical analysis indicates that the new approach converges globally to a first-order critical point under classical assumptions. In addition, the local convergence is also studied. Extensive numerical experiments for unconstrained optimization problems are reported.</div>
Unconstrained optimization, Trust-region framework, Nonmonotone technique, Theoretical convergence
15
33
http://ijmsi.ir/browse.php?a_code=A-10-3031-1&slc_lang=en&sid=1
M.
kimiaei
kimiaeim83@univie.ac.at
10031947532846008764
10031947532846008764
Yes
Vienna University
H.
esmaeili
esmaeili47@yahoo.com
10031947532846008765
10031947532846008765
No
Bu Ali University
F.
rahpeymaii
rahpeyma_83@yahoo.com
10031947532846008766
10031947532846008766
No
Payame Noor