In this paper, penalty parameter for linearly constrained 0-1 quadratic programming is improved.
本文改进了带线性约束0 - 1二次规划问题的罚参数下界。
This paper proposes a trust-region-type succesive linear algorithm with self-adjusted penalty parameter for nonlinear constrained optimization.
对于非线性约束最优化,提出了一个自动调节罚因子的信赖域类型的逐次线性算法。
To avoid reduction of critical time step size due to the penalty parameter for contact force calculation, a local density factor is introduced.
为避免用于计算接触力的罚函数法中的罚因子影响显式仿真算法的稳定性 ,采用了局部质量密度因子来抵消罚因子的负面影响。
This paper researches the parameters (kernel, penalty parameter c) of SVM and the dimension of feature, which influence aerial image segmentation and classification.
研究了支持向量机参数(核函数、惩罚因子c)和影像特征维数对航空影像分割与分类的影响。
First, the penalty finite element balance equation for quasi-static problem is obtained by the application of Galerkin weighted residual method and the introduction of penalty parameter.
首先,采用伽辽金加权余量法推导有限元平衡方程,利用罚方法得到求解拟静态问题的罚有限元公式。
Detailed convergence results for the dual algorithm are given and it is proved that there exists a threshold, and the algorithm will converge if the penalty parameter is less than the threshold.
对这一算法给出了精细的收敛性结果,证明了存在罚参数的一个阈值,当罚参数小于这个阈值时,算法收敛。
Detailed convergence results for the dual algorithm are given and it is proved that there exists a threshold, and the algorithm will converge if the penalty parameter is less than the threshold.
对这一算法给出了精细的收敛性结果,证明了存在罚参数的一个阈值,当罚参数小于这个阈值时,算法收敛。
应用推荐