基于稀疏鲁棒M-投资选择模型的鲁棒Half算法
发布时间：2019-01-03 20:58
【摘要】：为得到鲁棒、稀疏的投资组合,提出稀疏鲁棒M-投资选择模型,并且基于L1/2正则化理论和Half阈值算法,构建鲁棒Half阈值算法求解稀疏鲁棒M-投资选择问题.数值实验表明,该算法不仅比Lasso算法收敛速度更快,而且在期望值固定的情况下得到的风险更小、更平稳.
[Abstract]:In order to obtain a robust and sparse portfolio, a sparse robust M- investment selection model is proposed. Based on L1 / 2 regularization theory and Half threshold algorithm, a robust Half threshold algorithm is constructed to solve the sparse robust M- investment selection problem. Numerical experiments show that the proposed algorithm not only converges faster than the Lasso algorithm, but also has a lower and more stable risk when the expected value is fixed.
【作者单位】： 西安工程大学理学院;
【基金】：国家自然科学基金资助项目(11201362) 陕西省教育厅自然科学专项基金资助项目(14JK1305)
【分类号】：F224;F830.59
，
本文编号：2399861
[Abstract]:In order to obtain a robust and sparse portfolio, a sparse robust M- investment selection model is proposed. Based on L1 / 2 regularization theory and Half threshold algorithm, a robust Half threshold algorithm is constructed to solve the sparse robust M- investment selection problem. Numerical experiments show that the proposed algorithm not only converges faster than the Lasso algorithm, but also has a lower and more stable risk when the expected value is fixed.
【作者单位】： 西安工程大学理学院;
【基金】：国家自然科学基金资助项目(11201362) 陕西省教育厅自然科学专项基金资助项目(14JK1305)
【分类号】：F224;F830.59
，
本文编号：2399861
本文链接：https://www.wllwen.com/jingjifazhanlunwen/2399861.html
