高维约束矩阵回归问题

发布时间：2018-04-02 02:35

本文选题：矩阵回归　切入点：非凸约束　出处：《运筹学学报》2017年02期

【摘要】：高维约束矩阵回归是指高维情况下带非凸约束的多响应多预测统计回归问题,其数学模型是一个NP-难的矩阵优化,它在机器学习与人工智能、医学影像疾病诊疗、基因表达分析、脑神经网络、风险管理等领域有广泛应用.从高维约束矩阵回归的优化理论和算法两方面总结和评述这些新成果,同时,列出了相应的重要文献.
[Abstract]:High-dimensional constrained matrix regression is a multi-response and multi-predictive statistical regression problem with non-convex constraints under high-dimensional conditions. Its mathematical model is an NP-difficult matrix optimization, which is used in machine learning and artificial intelligence, medical imaging disease diagnosis and treatment. Gene expression analysis, neural network, risk management and other fields are widely used. These new achievements are summarized and reviewed from the optimization theory and algorithm of high dimensional constraint matrix regression. At the same time, some important literatures are listed.
【作者单位】：北京交通大学理学院;南安普顿大学数学科学学院;
【基金】：国家自然科学基金(Nos.11431002,11671029)
【分类号】：O212.1

【相似文献】