When dealing with a regular (fixed-support) one-parameter distribution, the corresponding maximum-likelihood estimator (MLE) is, to a good approximation, normally distributed. But, when the support boundaries are func...When dealing with a regular (fixed-support) one-parameter distribution, the corresponding maximum-likelihood estimator (MLE) is, to a good approximation, normally distributed. But, when the support boundaries are functions of the parameter, finding good approximation for the sampling distribution of MLE (needed to construct an accurate confidence interval for the parameter’s true value) may get very challenging. We demonstrate the nature of this problem, and show how to deal with it, by a detailed study of a specific situation. We also indicate several possible ways to bypass MLE by proposing alternate estimators;these, having relatively simple sampling distributions, then make constructing a confidence interval rather routine. .展开更多
When dealing with a regular (fixed-support) one-parameter distribution, the corresponding maximum-likelihood estimator (MLE) is, to a good approximation, normally distributed. But, when the support boundaries are func...When dealing with a regular (fixed-support) one-parameter distribution, the corresponding maximum-likelihood estimator (MLE) is, to a good approximation, normally distributed. But, when the support boundaries are functions of the parameter, finding good approximation for the sampling distribution of MLE (needed to construct an accurate confidence interval for the parameter’s true value) may get very challenging. We demonstrate the nature of this problem, and show how to deal with it, by a detailed study of a specific situation. We also indicate several possible ways to bypass MLE by proposing alternate estimators;these, having relatively simple sampling distributions, then make constructing a confidence interval rather routine. .展开更多
Deficiencies of the performance-based iterative learning control （ILC） for the non-regular systems are investigated in detail, then a faster control input updating and lifting technique is introduced in the design o...Deficiencies of the performance-based iterative learning control （ILC） for the non-regular systems are investigated in detail, then a faster control input updating and lifting technique is introduced in the design of performance index based ILCs for the partial non-regular systems. Two ldnds of optimal ILCs based on different performance indices are considered. Finally, simulation examples are given to illustrate the feasibility of the proposed learning controls.展开更多
A method to construct strongly non regular order bounded operators from a classical Banach lattice C into any separable Banach lattice F without Dedekind σ completeness is presented in this paper. A r...A method to construct strongly non regular order bounded operators from a classical Banach lattice C into any separable Banach lattice F without Dedekind σ completeness is presented in this paper. A result concerning the order bounded norm and the regular norm is also contained.展开更多
For any given 0 〈α 〈 β 〈 ∞, we construct a tree such that under tree metric, the Hausdorff dimension of the corresponding boundary is α, but both the Packing dimension and the boxing dimension are β. Applying ...For any given 0 〈α 〈 β 〈 ∞, we construct a tree such that under tree metric, the Hausdorff dimension of the corresponding boundary is α, but both the Packing dimension and the boxing dimension are β. Applying the connection between tree and iterated functions system, non- regular fractal sets on real line are constructed. Moreover, the method adopted in our paper is different from those which have been used before for constructing non-regular fractal set in general metric space.展开更多
针对合成孔径雷达(synthetic aperture radar,SAR)稀疏成像中目标反射率易低估、目标结构特征难以精确提取的问题,提出一种基于非凸和相对全变分(relative total variation,RTV)正则化的稀疏SAR成像算法。该算法利用非凸惩罚抑制偏差效...针对合成孔径雷达(synthetic aperture radar,SAR)稀疏成像中目标反射率易低估、目标结构特征难以精确提取的问题,提出一种基于非凸和相对全变分(relative total variation,RTV)正则化的稀疏SAR成像算法。该算法利用非凸惩罚抑制偏差效应、RTV自适应保护图像结构,在交替方向乘子法(alternating direction method of multipliers,ADMM)分布式优化框架下,实现多个正则项的协同优化增强。为更好地提高成像效率和降低内存占用量,利用匹配滤波(match filter,MF)算子构造测量矩阵进行近似观测,并对重建的SAR图像质量进行定量评价。仿真与实测数据处理结果表明,所提方法可有效抑制噪声杂波,在保证空间分辨率的情况下有效提高目标重建精度和辐射分辨率。展开更多
个性化联邦学习侧重于为各客户端提供个性化模型,旨在提高对异构数据的处理性能,然而现有的个性化联邦学习算法大多以增加客户端参数量为代价提高个性化模型的性能,使计算变得复杂.为了解决此问题,文中提出基于稀疏正则双层优化的个性...个性化联邦学习侧重于为各客户端提供个性化模型,旨在提高对异构数据的处理性能,然而现有的个性化联邦学习算法大多以增加客户端参数量为代价提高个性化模型的性能,使计算变得复杂.为了解决此问题,文中提出基于稀疏正则双层优化的个性化联邦学习算法(Personalized Federated Learning Based on Sparsity Regularized Bi-level Optimization,pFedSRB),在客户端的个性化更新中引入l 1范数稀疏正则化,提升个性化模型的稀疏度,避免不必要的客户端参数更新,降低模型复杂度.将个性化联邦学习建模为双层优化问题,内层优化采用交替方向乘子法,可提高学习速度.在4个联邦学习基准数据集上的实验表明,pFedSRB在异构数据上表现出色,在提高模型性能的同时有效降低训练用时和空间成本.展开更多
文摘When dealing with a regular (fixed-support) one-parameter distribution, the corresponding maximum-likelihood estimator (MLE) is, to a good approximation, normally distributed. But, when the support boundaries are functions of the parameter, finding good approximation for the sampling distribution of MLE (needed to construct an accurate confidence interval for the parameter’s true value) may get very challenging. We demonstrate the nature of this problem, and show how to deal with it, by a detailed study of a specific situation. We also indicate several possible ways to bypass MLE by proposing alternate estimators;these, having relatively simple sampling distributions, then make constructing a confidence interval rather routine. .
文摘When dealing with a regular (fixed-support) one-parameter distribution, the corresponding maximum-likelihood estimator (MLE) is, to a good approximation, normally distributed. But, when the support boundaries are functions of the parameter, finding good approximation for the sampling distribution of MLE (needed to construct an accurate confidence interval for the parameter’s true value) may get very challenging. We demonstrate the nature of this problem, and show how to deal with it, by a detailed study of a specific situation. We also indicate several possible ways to bypass MLE by proposing alternate estimators;these, having relatively simple sampling distributions, then make constructing a confidence interval rather routine. .
基金supported by the National Natural Science Foundation of China (No.60774023)Hunan Provincial Natural Science Foundation (No.06JJ50141)
文摘Deficiencies of the performance-based iterative learning control （ILC） for the non-regular systems are investigated in detail, then a faster control input updating and lifting technique is introduced in the design of performance index based ILCs for the partial non-regular systems. Two ldnds of optimal ILCs based on different performance indices are considered. Finally, simulation examples are given to illustrate the feasibility of the proposed learning controls.
文摘A method to construct strongly non regular order bounded operators from a classical Banach lattice C into any separable Banach lattice F without Dedekind σ completeness is presented in this paper. A result concerning the order bounded norm and the regular norm is also contained.
文摘随着高分辨率对地观测要求的不断提高,合成孔径雷达(Synthetic Aperture Radar,SAR)的应用将越来越广泛。针对高分辨率SAR成像存在数据量大、存储难度高、计算时间长等问题,目前常用的解决方法是在SAR成像模型中引入压缩感知(Compressed Sensing,CS)的方法降低采样率和数据量。通常使用单一的正则化作为约束条件,可以抑制点目标旁瓣,实现点目标特征增强,但是观测场景中可能存在多种目标类型,因此使用单一正则化约束难以满足多种特征增强的要求。本文提出了一种基于复合正则化的稀疏高分辨SAR成像方法,通过压缩感知降低数据量,并使用多种正则化的线性组合作为约束条件,增强观测场景中不同类型目标的特征,实现复杂场景中高分辨率对地观测的要求。该方法在稀疏SAR成像模型中引入非凸正则化和全变分(Total Variation,TV)正则化作为约束条件,减小稀疏重构误差、增强区域目标的特征,降低噪声对成像结果的影响,提高成像质量;采用改进的交替方向乘子法(Alternating Direction Method of Multipliers,ADMM)实现复合正则化约束的求解,减少计算时间、快速重构图像;使用方位距离解耦算子代替观测矩阵及其共轭转置,进一步降低计算复杂度。仿真和实测数据实验表明,本文所提算法可以对点目标和区域目标进行特征增强,减小计算复杂度,提高收敛性能,实现快速高分辨的图像重构。
文摘For any given 0 〈α 〈 β 〈 ∞, we construct a tree such that under tree metric, the Hausdorff dimension of the corresponding boundary is α, but both the Packing dimension and the boxing dimension are β. Applying the connection between tree and iterated functions system, non- regular fractal sets on real line are constructed. Moreover, the method adopted in our paper is different from those which have been used before for constructing non-regular fractal set in general metric space.
文摘针对合成孔径雷达(synthetic aperture radar,SAR)稀疏成像中目标反射率易低估、目标结构特征难以精确提取的问题,提出一种基于非凸和相对全变分(relative total variation,RTV)正则化的稀疏SAR成像算法。该算法利用非凸惩罚抑制偏差效应、RTV自适应保护图像结构,在交替方向乘子法(alternating direction method of multipliers,ADMM)分布式优化框架下,实现多个正则项的协同优化增强。为更好地提高成像效率和降低内存占用量,利用匹配滤波(match filter,MF)算子构造测量矩阵进行近似观测,并对重建的SAR图像质量进行定量评价。仿真与实测数据处理结果表明,所提方法可有效抑制噪声杂波,在保证空间分辨率的情况下有效提高目标重建精度和辐射分辨率。
文摘个性化联邦学习侧重于为各客户端提供个性化模型,旨在提高对异构数据的处理性能,然而现有的个性化联邦学习算法大多以增加客户端参数量为代价提高个性化模型的性能,使计算变得复杂.为了解决此问题,文中提出基于稀疏正则双层优化的个性化联邦学习算法(Personalized Federated Learning Based on Sparsity Regularized Bi-level Optimization,pFedSRB),在客户端的个性化更新中引入l 1范数稀疏正则化,提升个性化模型的稀疏度,避免不必要的客户端参数更新,降低模型复杂度.将个性化联邦学习建模为双层优化问题,内层优化采用交替方向乘子法,可提高学习速度.在4个联邦学习基准数据集上的实验表明,pFedSRB在异构数据上表现出色,在提高模型性能的同时有效降低训练用时和空间成本.
