学术动态

位置: 首页 > 科学研究 > 学术动态 > 正文

学术报告3: Dijana Mosic — Composite outer inverses for rectangular matrices

时间:2023-01-04 作者: 点击数:

报告时间202316日(星期五)1400-15:30

报告地点腾讯会议:796-160-323

Dijana Mosic 教授

工作单位University of Nis

举办单位:万象城官方网站(中国)有限公司

报告人简介:Dijana MosicProfessor of University of NisShe has published more than 165 scientific papers in mathematics. These papers are cited more than 800 times. She is the author of one book of problems, coauthor of one book of problems and author of one monograph on generalized inverses. She is an excellent mathematican. She won the Award “For women in Science”, LOREAL-UNESCO in 2012,and Award for achievements in Mathematical Sciences, Mathematical Society of Serbia, 2018.

报告简介 Various compositions of the Drazin inverse, the group inverse or the core-EP inverse with the Moore-Penrose inverse have investigated last years. Solving some type of matrix equations, we introduce three new generalized inverses of a rectangular matrix, which are called the OMP, MPO and MPOMP inverses, because the outer inverse and the Moore-Penrose inverse are incorporated in their definition. As a consequence, the notion of DMP, MPD, CMP and MPCEP inverses for a square matrix are covered by one general definition and extended to a rectangular matrix. We propose a common term, composite outer inverses, to denote such compositions of outer inverses and the Moore-Penrose inverse. Characterizations of the OMP, MPO and MPOMP inverses are derived as well as some properties of projectors determined by these new inverses. We establish maximal classes of matrices for which the representations of composite outer inverses are valid. Also, the integral and limit representations for OMP, MPO and MPOMP inverses are investigated. Possible applications of composite outer inverses are given too and interesting topics for further research are considered.


上一篇:学术报告4:储德林 — Alternating Nonnegative Least Squares for Nonnegative Matrix Factorization (2)

下一篇:学术报告2: 储德林 — Alternating Nonnegative Least Squares for Nonnegative Matrix Factorization (I)