解线性最小二乘问题的2个混合交替CQ算法  

Two Hybrid Alternating CQ-algorithms for Solving the Linear Least Square Problem

作　　者：付元敏 朱立军 贺龙 FU Yuanmin;ZHU Lijun;HE Long(School of Mathematics and Information Science,North Minzu University,Yinchuan 750021,Ningxia)

机构地区：[1]北方民族大学数学与信息科学学院,宁夏银川750021

出　　处：《四川师范大学学报（自然科学版）》2020年第2期202-211,共10页Journal of Sichuan Normal University（Natural Science）

基　　金：国家自然科学基金(61362033)、宁夏自然科学基金(NZ17015)和教育部科技发展中心基金(2018A01018)。

摘　　要：给出2个改进的混合交替CQ算法求解线性最小二乘问题.许多学者研究这个问题并提出了各种各样的算法去解决它.混合交替CQ算法在科学领域有着广泛的应用,它起源于相位恢复、医学图像重建、强度可调放射疗法、信号工程和X光断层摄影技术.给出2个新的算法的弱收敛性证明,作为应用考虑将其求解线性最小二乘问题.In this paper,we present two modified hybrid alternating CQ-algorithms to study the linear least square problem.Many researchers studied this problem and proposed various algorithms to solve it.The hybrid alternating CQ-algorithms,arising from phase retrievals,medical image reconstruction,modeling of intensity modulated radiation therapy,signal processing and tomography,have broad application in many scientific areas.We give two new algorithms and give the proof of their weak convergence.As application,we consider the linear least square problem.

关 键 词：分裂可行性问题 弱收敛性 混合交替CQ算法 线性最小二乘问题 

分 类 号：O177[理学—数学]