Please use this identifier to cite or link to this item:
http://acervodigital.unesp.br/handle/11449/9651
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Teixeira, MCM | - |
dc.contributor.author | Zak, S. H. | - |
dc.date.accessioned | 2014-05-20T13:28:54Z | - |
dc.date.accessioned | 2016-10-25T16:48:24Z | - |
dc.date.available | 2014-05-20T13:28:54Z | - |
dc.date.available | 2016-10-25T16:48:24Z | - |
dc.date.issued | 1998-07-01 | - |
dc.identifier | http://dx.doi.org/10.1109/72.701176 | - |
dc.identifier.citation | IEEE Transactions on Neural Networks. New York: IEEE-Inst Electrical Electronics Engineers Inc., v. 9, n. 4, p. 629-638, 1998. | - |
dc.identifier.issn | 1045-9227 | - |
dc.identifier.uri | http://hdl.handle.net/11449/9651 | - |
dc.identifier.uri | http://acervodigital.unesp.br/handle/11449/9651 | - |
dc.description.abstract | Continuous-time neural networks for solving convex nonlinear unconstrained;programming problems without using gradient information of the objective function are proposed and analyzed. Thus, the proposed networks are nonderivative optimizers. First, networks for optimizing objective functions of one variable are discussed. Then, an existing one-dimensional optimizer is analyzed, and a new line search optimizer is proposed. It is shown that the proposed optimizer network is robust in the sense that it has disturbance rejection property. The network can be implemented easily in hardware using standard circuit elements. The one-dimensional net is used as a building block in multidimensional networks for optimizing objective functions of several variables. The multidimensional nets implement a continuous version of the coordinate descent method. | en |
dc.format.extent | 629-638 | - |
dc.language.iso | eng | - |
dc.publisher | Institute of Electrical and Electronics Engineers (IEEE) | - |
dc.source | Web of Science | - |
dc.subject | analog networks | pt |
dc.subject | coordinate descent | pt |
dc.subject | derivative free optimization | pt |
dc.subject | unconstrained optimization | pt |
dc.title | Analog neural nonderivative optimizers | en |
dc.type | outro | - |
dc.contributor.institution | Universidade Estadual Paulista (UNESP) | - |
dc.contributor.institution | Purdue Univ | - |
dc.description.affiliation | UNESP, Dept Elect Engn, FEIS, BR-15385000 Ilha Solteira, SP, Brazil | - |
dc.description.affiliation | Purdue Univ, Sch Elect & Comp Engn, W Lafayette, IN 47907 USA | - |
dc.description.affiliationUnesp | UNESP, Dept Elect Engn, FEIS, BR-15385000 Ilha Solteira, SP, Brazil | - |
dc.identifier.doi | 10.1109/72.701176 | - |
dc.identifier.wos | WOS:000074419800005 | - |
dc.rights.accessRights | Acesso restrito | - |
dc.relation.ispartof | IEEE Transactions on Neural Networks | - |
Appears in Collections: | Artigos, TCCs, Teses e Dissertações da Unesp |
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.