通过引入正则化方法来改善解的稳定性。
Thestability of the solution is improved by the Tiknonov's regularization method.
贝叶斯正则化方法提高BP神经网络的泛化能力。
Bayes' regularization raises the ability to extend of BP neural network.
对现有的各向异性扩散的正则化方法进行了分析。
The existing regularity method about anisotropic diffusion is introduced.
恢复的方法有正则化方法、迭代方法、统计方法等。
The usual methods include regularization method, iterative method, stochastic method and so on.
第二章主要介绍反演问题的正则化方法和最优化方法。
In Chapter 2, we concentrate on the regularization methods and optimization methods for inverse problems.
重力场球谐函数系数可以通过正则化方法由重力梯度算出。
The spherical harmonic coefficients of the gravity field can be obtained from gravity gradients through regularization scheme.
该方法基于非线性积分方程，用正则化方法求出电导率分布。
The method, based on nonlinear integral equation, is used to solve for the formation conductivity profile by using regularization.
第一章为绪论，简单描述了熵正则化方法与罚函数法的研究现状；
The first chapter is the introduction for the entropy regularization method and penalty function method.
论述了由单幅X射线投影进行轴对称物体密度重建的正则化方法。
The regularization method for density reconstruction of axially symmetric objects using a single Xray projection is described.
用抛物正则化方法和弱收敛技巧，证明了最优控制的存在性和稳定性。
The existence and stability of the optimal control is established based on the parabolic regularization method and the weak convergence technique.
首先，提出了一种基于改进正则化方法的有限角度CT图像重建算法。
Firstly, a modified Tikhonov regularization ct image reconstruction algorithm from limited-angle is proposed.
本文采用的是辅助边界条件方法(ABC方法)，或者非局部正则化方法。
The Auxiliary Boundary Condition method (ABC method) or non-local regularization method is used in this thesis.
本文旨在研究获得两个逆热传导问题稳定解的正则化方法及其数值实现问题。
The aim of this thesis is to study the Regularization Method for stable solution of two inverse heat conduction problems and study their numerical implements.
我们将利用正则化方法和上下解技巧给出局部古典解和整体古典解的存在唯一性。
We will use regularization method and upper and lower solution technique to give the local existence, global existence and uniqueness results.
第一章介绍了反问题的基本概念及其应用领域，并总结了一些基本的正则化方法。
In the first chapter, we present the basic notion and some applications of inverse problems. We also summarize some important regularization methods.
正则化方法是提高SAR图像分辨率的有效方法，但传统的迭代算法计算速度较慢。
Regularization method is effective to improve the resolution of SAR image, but the traditional iterative method for resolving the regularization model runs slow.
在理想状态下，假定火源为线火源，采用离散正则化方法对矿井隐蔽线火源进行反演。
Under the perfect condition, supposes fire as line fire, applying discrete regularization method into the inversion of coal mine concealed fi.
运用正则化方法和上下解技巧证明了上述问题的古典正解的局部存在性及其可延拓性。
The method of regularization and the technique of upper and lower solutions are employed to show the local existence and the continuation of the positive classical solution of the above problem.
为了获得稳定而满意的解，我们采用直方图约束下的正则化方法对连续近似迭代进行约束。
To obtain a stable solution, in our method, successive approximation process is constrained by prior histogram and laplacian regularization.
数值实验表明，径向基函数配置点方法与正则化方法耦合能有效求解椭圆型偏微分方程反问题。
It is concluded that the radial basis function collocation techniques coupled with regularization methods could be competitive alternatives to existing methods for these problems.
通过高精度的数控移动工件台获取密集的样本数据，并在神经网络训练过程中采用贝叶斯正则化方法。
Dense sample data are acquired by using numerical control platform of high precision, and the Bayesian generalization is adopted during training the neural network.
提出一种新的非线性结构张量计算方法，扩展了基于迹的PDE正则化方法，使其适用于矩阵值数据场；
A new nonlinear structure tensor calculation method is presented. We extend the trace-based PDE regularization method to matrix-valued data fields.
鉴于此，必须采用正则化方法，本文中选用的是截断奇异值分解，其正则化参数用l -曲线准则来确定。
In this thesis, we choose truncated singular value decomposition to solve the resulting matrix equations, while the regularization parameter of TSVD is determined by the L-curve criterion.
本文是在MPI网络并行环境中，将求解二维热物性方程参数的反问题用正则化方法结合并行遗传算法进行数值求解。
This thesis focuses on determination of parameters in a two-dimension heat conduction equation on the MPI network parallel environment, by solving an inverse problem using the regularization method.
基于相空间重构的非线性预报思想，建立一个时滞的BP神经网络模型，采用贝叶斯正则化方法提高BP网络的泛化能力。
Based on nonlinear prediction ideas of reconstructing phase space, this paper presents a time delay BP neural network model, whose generalization is improved utilizing Bayes' regularization.
因此许多学者提出了各种求解反问题的方法，比如脉冲谱方法，最佳摄动量法，蒙特卡罗方法，各种优化方法和正则化方法等。
So various methods are proposed by scholars to solve these problems, such as pulse spectrum method, the best perturbation method, Monte Carlo method, optimized and regularization method.
本文运用正则化方法证明了一类退化抛物方程解的存在唯一性，讨论了解的全局存在性与爆破，并在一定的初值条件下得到了解的爆破速率。
In this paper, we establish the local existence and uniqueness of the solution by using regularization method. We also obtain the global existence and nonexistence. Finally, we get the blow-up rate.
本文根据常微分方程参数反问题的数学理论，将正交化方法同有限差分法结合用于确定水质模型参数，并与正则化方法、最速下降法和共轭梯度法作了比较。
The comparison of the calculation results show that orthogonal rule method is fast, simple and reliable, and is applicable to the calculation of the water quality modeling parameters.
本文根据常微分方程参数反问题的数学理论，将正交化方法同有限差分法结合用于确定水质模型参数，并与正则化方法、最速下降法和共轭梯度法作了比较。
The comparison of the calculation results show that orthogonal rule method is fast, simple and reliable, and is applicable to the calculation of the water quality modeling parameters.
应用推荐