查询词典 gradient method

Two ways are used to design the network,the one is the direct energy descent method,and the other is the gradient descent method .

有两种方法可以用来设计网络，一是能量下降法，另一个是梯度下降法。

Moreover, a new fast learning method of fuzzy systems both are based on genetic algorithms and gradient descent method is proposed.

引 言最近几年来，诸如模糊推理、模糊建模、模糊逻辑控制器等模糊系统得到了广泛的研究和应用。

Currently, the learning algorithms for training Feedforward Multilayer Neural Network can be classified into two categories: gradient descent method and layer by layer optimization method.

目前前向神经网络的学习算法主要分为两类：梯度下降算法和逐层优化算法。

In this paper, a new edge detection method——method using extremum of first directional derivative is proposed, based upon a general gradient edge model.

文中通过傅里叶变换，在频率域，通过求垂直于边缘的一阶方向导数，并检测其极值点来确定边缘。

The Gradient—Hamilton function method is presented first. Its formula with multiplier function and relax factors quated in solving constrained shape optimization problems is derived. The method is carried out on micro computer to solve shape optimal design of column.

提出了梯度—H函数极值法，并推导出了求解约束形状优化问题时引入的乘子函数和松驰因子的具体公式，在微机上实现了这种方法，并用其解决了立柱的形状优化问题。

The main results of the investigation are as following:(1) Some important criteria for the design of multi-layer structure are acquired: the ratio of the layer thickness to the contact diameter should be controlled to a value about 0.8; increasing the number of the gradient layers and elasticity modulus of the top layer would benefit to the yield strength and the stress status at the interface.（2）By means of molecular dynamics simulations for a indentation and sliding process, it is discovered that dislocations of 45 direction to the surface occur in the surfaces layer, which is correspond to the direction of the plastic flows observed in the macroscopic continuous medium. The irreversible deformation caused by the dislocation and the atom piles at the surface are responsible for the irreversibility of the force response. They are also the main factors for the friction generation. The temperature and the indentation velocity have significant effects on the deformation behaviors of the material of the surface layer. The yield strength would increase obviously if the indentation velocity is greater than 10 m/s.（3）A method to evaluate the fracture toughness of coating by estimating the crack-density functionβof a Wicker''s indentation was proposed. Meanwhile a new surface treatment technology was developed using non-equilibrium plasma.（4）A MD-FE hybrid method and corresponding software was developed to analysis the tribological behaviors for multiplayer system.

研究主要内容和结果有：（1）得到指导梯度膜设计的几个准则：控制膜层厚度与接触半径比在0.8左右为宜；适当增加梯度层层数和顶层弹性模量，有利于提高抗屈服能力和改善界面处的应力状态；（2）通过对嵌入和滑动摩擦过程的分子动力学模拟，发现表层产生一个与宏观连续介质的塑性流动方向观测一致的45 方向位错；位错造成的不可逆变形和表面的原子堆积，是力的响应不可逆性的原因，也是构成摩擦力的主要因素；温度和嵌入速度对材料变形特性有显著的影响，当嵌入速度大于10 m/s时，屈服强度明显增加；（3）提出压痕裂纹密度系数β半定量判定膜层断裂韧性的方法；并开发出表面强化新工艺；（4）开发出分析表面摩擦学特性的MD-FE复合模拟方法和相应模拟软件。

The OTSU algorithm （maximization of interclass variance） is considered as one of the superior threshold selection methods.But this method can not segment the path mark image successfully for our indigenous design and manufactured vision navigation intelligent vehicle under bright illumination.A new method based on regional Roberts gradient is proposed in order to solve this problem.

OTSU法被认为是图像分割中阈值自动选取的最优方法之一，针对自主开发的视觉导航区域交通智能车辆的导航路径在强光照条件下不能直接应用此方法取得准确分割效果的问题，提出了一种基于区域的罗伯特梯度算子的图像分割方法。

Two special cases are discussed: flow due to a constant pressure gradient and flow due to a periodic one; In chapter 4, on the basis of the method of fractional element proposed by Schiessel et al and of the generalized fractional element networks proposed by Xu and Tan and by the use of discretization method for solving the inverse Laplace transform, we obtain the creep compliance of Zener and Poynting-Thomson models which are composed of the generalized fractional element networks.

并讨论了如下两种特殊情形：常压力梯度下的流动和周期性压力梯度下的流动；第四章在Schiessel等提出的分数阶单元法和徐明瑜等提出的广义分数阶单元网络的基础上，应用离散求逆Laplace变换的方法，给出并讨论了广义分数阶单元网络Zener和Poynting-Thomson模型的蠕变柔量。

The nonlinearity and discrete gradient inherited in CAViaR model is a conundrum for parameter estimation. We take the asymmetric Laplace distribution with scale parameter as the error process; indicate the variance has a minimum positive value when the scale parameter is a constant, conflicting with the distribution of real financial data. Further we estimate the parameters of indirect TARCH-CAViaR model base on Bayesian framework and Markov chain Monte Carlo method. The optimal scale parameter can also be obtained by Markov chain Monte Carlo method.

CAViaR一般模型中递归分位回归方程的非线性和非连续可微性是参数估计的一个难题，基于含有尺度参数的不对称拉普拉斯分布作为误差过程，指出将尺度参数固定为常数会导致不对称拉普拉斯分布随机变量的方差存在最小正值的限制，与实际金融数据分布不符；进而提出采用贝叶斯分析和马尔科夫链蒙特卡罗模拟方法，估计间接TARCH-CAViaR模型的参数，并可获得尺度参数的合理估计。

An applied method was put forward to select regularization parameters in the iterative process of gradient regularization method.

并给出了该方法在计算过程中求正则解时一种实用正则参数选取方法。

I can not make it blossom and suits me

我不能让树为我开花

When temperatures are above approximately 80 °C discolouration of the raceways or rolling elements is a frequent feature.

当温度高于 80 °C 左右时，滚道或滚动元件褪色是很常见的特征。