https://scholars.lib.ntu.edu.tw/handle/123456789/490445
標題: | Convergence of the Exponentiated Gradient Method with Armijo Line Search | 作者: | YEN-HUAN LI Cevher, Volkan |
關鍵字: | Armijo line search; Exponentiated gradient method; Peierls–Bogoliubov inequality; Self-concordant likeness | 公開日期: | 2019 | 卷: | 181 | 期: | 2 | 起(迄)頁: | 588-607 | 來源出版物: | Journal of Optimization Theory and Applications | 摘要: | Consider the problem of minimizing a convex differentiable function on the probability simplex, spectrahedron, or set of quantum density matrices. We prove that the exponentiated gradient method with Armijo line search always converges to the optimum, if the sequence of the iterates possesses a strictly positive limit point (element-wise for the vector case, and with respect to the Löwner partial ordering for the matrix case). To the best of our knowledge, this is the first convergence result for a mirror descent-type method that only requires differentiability. The proof exploits self-concordant likeness of the log-partition function, which is of independent interest. © 2018, Springer Science+Business Media, LLC, part of Springer Nature. |
URI: | https://scholars.lib.ntu.edu.tw/handle/123456789/490445 https://www.scopus.com/inward/record.uri?eid=2-s2.0-85058053363&doi=10.1007%2fs10957-018-1428-9&partnerID=40&md5=ad0a273b6464d403e7ae8ae9af8514bc |
ISSN: | 00223239 | DOI: | 10.1007/s10957-018-1428-9 |
顯示於: | 資訊工程學系 |
在 IR 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。