助手标题  
全文文献 工具书 数字 学术定义 翻译助手 学术趋势 更多
查询帮助
意见反馈
   local minima 的翻译结果: 查询用时:0.088秒
图标索引 在分类学科中查询
所有学科
自动化技术
数学
地球物理学
工业通用技术及设备
更多类别查询

图标索引 历史查询
 

local minima     
相关语句
  局部极小
     Based on probabilistic neural networks (PNN), a new way of fault diagnosis of steam turbine units is suggested to overcome problems met with back propagation neural networks (BPNN) like slow convergence of learning and liability of dropping into local minima.
     针对反向传播神经网络(BPNN)学习收敛速度慢、易陷入局部极小值等问题,提出了采用概率神经网络(PNN)的汽轮发电机组故障诊断方法。
短句来源
     The paper advanced a new way of fusion modeling by means of the use of mutative scale chaos optimization method and BP networks to overcome problems in makeup parameters,convergence speed and local minima of BP networks.
     针对 BP神经网络存在的网络结构参数、收敛速度、局部极小等问题 ,提出基于变尺度混沌遗传算法 ( MSCGA)与 BP神经网络相互结合的混合建模新方法 ,可对 BP网络隐层节点数、网络权值、阈值等结构参数进行快速优化设计。
短句来源
     The deficiencies of gradient descent method include the slow speed of convergence, the problem of local minima and the great influence of initial parameters on the performance of the network.
     在使用基于梯度下降的径向基函数(RBF)神经网络学习方法时,由于网络目标函数误差曲面极其复杂,因而产生了网络收敛速度慢,且容易陷入局部极小,网络初始值的设置对网络训练结果影响很大等问题。
短句来源
     Aimed at particle swarm optimization(PSO) algorithm being easily trapped into local minima value in multimodal function,a rotating surface transformation(RST) method was proposed.
     针对粒子群优化算法(PSO)应用于多极值点函数易陷入局部极小值,提出旋转曲面变换(RST)方法.
短句来源
     Furthermore, evolving algorithms, which can escape from local minima and get more suitable RNN models for the proposed problems, are more feasible for training various RNN structures.
     而进化算法相比于常规的梯度下降算法 ,用于训练不同的回归网络结构通用性好 ,且训练过程不受局部极小问题的困扰 ,适当规模的训练过程可以获得性能良好的神经网络模型。
短句来源
更多       
  局部最小
     Considering that consistent continuous Hopfield neural network lead to local minima or illegal results,the energy analysis of inconsistent continuous Hopfield neural network was given and the conditions under which network energy can increase,decrease or maintain were got.
     针对一致连续Hopfield网络容易陷入局部最小和不合理解的现象,给出了不一致连续Hopfield网络的能量分析,得到了使网络能量函数上升、下降和不变的条件.
短句来源
     The method solves the problem of local minima of ICP algorithm,and the described method is also useful in the field of comparative analysis for precision measurement.
     采用两步法配准克服了标准 ICP算法难以解决的局部最小问题 ,本算法也可广泛应用于精密测量时测量结果的比较分析。
短句来源
     Due to the localized nature, RBFN can be trained quickly and can avoid falling into local minima.
     由于其局部性质 ,RBFN能被快速训练 ,避免陷入局部最小
短句来源
     Genetic algorithm is a overall algorithm,it can escape local minima.
     遗传算法是全局算法,它能避免陷入局部最小
短句来源
     The discrete gradient algorithms and the algorithms of approximating the gradient with overlap vectors improve greatly in time complexity ,but the algorithms tend to stuck in local minima without getting the final result .
     离散梯度下降算法、交叠向量算法在运算时间上有了很大的改善,但是算法容易陷入局部最小值而得不到最终的解;
短句来源
更多       
  局部极小的
     By combining genetic algorithm with error back propagation neural network,a hybrid algorithm that trained neural network weight by real-coded adaptive mutation genetic algorithm is presented,and it overcomes the disadvantage that traditional neural network is easy to fall into local minima.
     通过遗传算法(GeneticAlgorithm)与误差反向传播(ErrorBackPropagation)神经网络相结合,提出了用实数编码的自适应变异遗传算法训练神经网络权重的混合算法,避免了传统神经网络易陷入局部极小的缺点.
短句来源
     To solve the slow convergence of BP algorithm and avoid getting stick in local minima, a general and simple training algorithm for feed-forward neural networks(FNN), named IBPM, which combines a class of nonlinear property of momentum term with the gradient descent of BP algorithm, is presented.
     针对BP算法收敛缓慢和易于陷入局部极小的缺点,将基于一类非线性特性的动量项引入BP算法的梯度搜索,提出前向神经网络(FNN)的一种通用且简单的全局训练算法(IBPM 算法)。
短句来源
     Protein is a flexible macromolecule and there are astronomical numbers of local minima on its energy hyper-surface. How to surmount the multiple-minima problem is the key in the prediction of protein structure.
     蛋白质是一个强柔性大分子的复杂体系,它的能量超表面存在着大量的局部极小,如何避免陷入局部极小的陷阱,找到全局能量极小构象成为解决蛋白质结构预测问题的关键。
     Compared with the conventional Hopfield neural network (HNN),TCNN would not be stuck into local minima by introducing chaos,which is generated by negative self feedback into HNN.
     该方法利用神经元的自反馈产生混沌动态 ,由于混沌动态特性具有很强的搜索全局最优解的能力 ,有效地避免了传统 Hopfield神经网络 ( HNN)极易陷入局部极小的缺陷 ;
短句来源
     Experiments of computer simulations show that the speed and precision of convergence and the ability of eluding local minima are far better than those of the conventional BP algorithm.
     仿真实验结果表明:SASSFBP算法在收敛速度与运算精度,以及避免陷入局部极小的能力等方面均明显优于常规的BP算法。
短句来源
更多       
  局部极小值
     Based on probabilistic neural networks (PNN), a new way of fault diagnosis of steam turbine units is suggested to overcome problems met with back propagation neural networks (BPNN) like slow convergence of learning and liability of dropping into local minima.
     针对反向传播神经网络(BPNN)学习收敛速度慢、易陷入局部极小值等问题,提出了采用概率神经网络(PNN)的汽轮发电机组故障诊断方法。
短句来源
     Aimed at particle swarm optimization(PSO) algorithm being easily trapped into local minima value in multimodal function,a rotating surface transformation(RST) method was proposed.
     针对粒子群优化算法(PSO)应用于多极值点函数易陷入局部极小值,提出旋转曲面变换(RST)方法.
短句来源
     In this paper, the principle and the calculation method of eigenfrequency analysis of structures with BEM are introduced. The approach formulas are given for the calculation of Bessel functions in the coefficient matrix of governing equations. The eigenfrequency of structures is obtained as such values that the complex determinant achieves local minima.
     本文阐述了用边界单元法计算结构自振频率与振型的原理与具体实施方法,给出了计算支配方程系数矩阵元素用的特殊函数的逼近公式,用求复行列式模的局部极小值方法计算各阶自振频率,用最小二乘法计算相应的各阶振型.
短句来源
     By accepting the new solution, whose quality is not better than the previous one, according to the different rules, the above two algorithms are possible to let the search escape from local minima and thus effectively solve the equal disks packing problem.
     这两个方法通过在搜索过程中以不同的规则接受不优于当前解的新解,使搜索有可能逃离局部极小值陷阱,从而有效地求解了等圆packing问题。
短句来源
     Especially, in the applications of optimization, neural control, signal processing and pattern recognition, the global asymptotic stability has become the core problem in the research in order to avoid the spurious response or local minima.
     特别地,在最优化、神经控制、信号处理和模式区别的应用中,须要求网络避免伪响应或发生局部极小值,因此,ANN网络的全局渐近稳定性是研究的核心问题。
短句来源
更多       

 

查询“local minima”译词为其他词的双语例句

 

查询“local minima”译词为用户自定义的双语例句

    我想查看译文中含有:的双语例句
例句
为了更好的帮助您理解掌握查询词或其译词在地道英语中的实际用法,我们为您准备了出自英文原文的大量英语例句,供您参考。
  local minima
It was proved that the conformational state that is close to the structure of the ADA complex with 1ADD known from X-ray study corresponds to one of the local minima of the potential surface.
      
The potential energy surface of all four complexes was found to contain local minima corresponding to a stable state with the η1-coordination of NO through the N atom and to two metastable isomers with the η1-O and η2-NO coordination.
      
The potential energy surface of all four complexes was found to contain local minima corresponding to a stable state with the η1-coordination of NO through the N atom and to two metastable isomers with the η1-O and η2-NO coordination.
      
The local minima corresponding to metastable isomers with a linear NO coordination through the oxygen atom and with a side η2 NO coordination were found on the potential energy surfaces of these compounds.
      
This behavior is qualitatively explained in terms of migration of excitons between local minima of the random potential in the plane of the quantum well.
      
更多          


In many papers of adaptive line enhancer (ALE) it was assumedthat the adaptive weights were uncorrelated with each other, and independent withinput signal. This leads the adaptive weight vector converges to the solution ofWiener-Hopf matrix equation by using widrow-Hoff least mean square error algo-rithm. In practice, it is found that the convergent weight vector is different fromthe solution of Weiner-Hopf matrix equation. While signal-to-noise ratio is smalland the weights is large, the difference is serious....

In many papers of adaptive line enhancer (ALE) it was assumedthat the adaptive weights were uncorrelated with each other, and independent withinput signal. This leads the adaptive weight vector converges to the solution ofWiener-Hopf matrix equation by using widrow-Hoff least mean square error algo-rithm. In practice, it is found that the convergent weight vector is different fromthe solution of Weiner-Hopf matrix equation. While signal-to-noise ratio is smalland the weights is large, the difference is serious. It's because the error surfaceof the adaptive iteration is not an ideal quadratic surface, and the error has localminima with respect to weights. In the process of iteration, the prediction errorremains at local minima of the error surface, so it doesn' t converge to the uniqueminimum point. In order to improve the performance of the ALE in the conditionof small SNR and large weights, an adaptive line enhancer with two stages ofweights is proposed. The ALE with large weights is changed into two stage ofsmaller weights. For example, an ALE with 1024 weights is changed to an ALEwith two stages of 32 weights. Since the later is smaller in weights, the meangradient of the error surface is large, and the convergence point of iteration maymove across the local minima to unique minimum of the error surface easily. Fina-lly, as an example, the sinusoid detection of the uncorrelated noise is discussed,and proved itself has the advantages over ordinary ALE.

为改善小信噪比输入和大权重数条件下的自适应线谱增强器(ALE)性能,本文提出双重权ALE,把大权重数的ALE化成两级小权重数的ALE。例如,可把1024个权重的ALE化成两级32个权重的ALE。由于每一级自适应这代均采用比较少的权重数,保证了误差平面具有较大的平均梯度,使收敛权重跳过误差平面上的局部极小点而达到最小点。最后,以非相关噪声中正弦波的检测为例,讨论双重权ALE的特性并进行计算机模拟实验。结果表明,双重权ALE较一般ALE有明显的优越性。

In this paper, the principle and the calculation method of eigenfrequency analysis of structures with BEM are introduced. The approach formulas are given for the calculation of Bessel functions in the coefficient matrix of governing equations. The eigenfrequency of structures is obtained as such values that the complex determinant achieves local minima. The least square method is used to calculate the eigenmodes of structures. Some examples show that the present method is an efficient one to solve the...

In this paper, the principle and the calculation method of eigenfrequency analysis of structures with BEM are introduced. The approach formulas are given for the calculation of Bessel functions in the coefficient matrix of governing equations. The eigenfrequency of structures is obtained as such values that the complex determinant achieves local minima. The least square method is used to calculate the eigenmodes of structures. Some examples show that the present method is an efficient one to solve the eigenvalue problems of structures.

本文阐述了用边界单元法计算结构自振频率与振型的原理与具体实施方法,给出了计算支配方程系数矩阵元素用的特殊函数的逼近公式,用求复行列式模的局部极小值方法计算各阶自振频率,用最小二乘法计算相应的各阶振型.计算实例表明,用边界单元法计算结构自振特性,前处理简单,精度较高,不失为一种有效的方法.

The neural net model, the feature extraction and the learning algorithm for the speech recognition by neural nets are investigated in this paper. A non-uniformly windowed pyramidical architecture is proposed, the normalizations of the feature vectors of two extraction algorithms are investgated and the Polak-Ribiere learning algorithm is modified to guarantee the weight vector not to converge to non-stable local minima. Speaker-dependent experments show that over 99% accuracy can be achieved for recognizing...

The neural net model, the feature extraction and the learning algorithm for the speech recognition by neural nets are investigated in this paper. A non-uniformly windowed pyramidical architecture is proposed, the normalizations of the feature vectors of two extraction algorithms are investgated and the Polak-Ribiere learning algorithm is modified to guarantee the weight vector not to converge to non-stable local minima. Speaker-dependent experments show that over 99% accuracy can be achieved for recognizing five vowels and ten digits. Speaker-independent experiments involved 90 speakers from 25 provinces of China show that 90.3% accuracy can be obtained for the digit recognition, and the performance can be increased by 1.0% after the pre-processing window is employed. The windowed model is more applicable for speech recognitions.

本文对神经网络法语音识别中的网络结构、特征提取以及学习算法进行了初步的研究.文中提出了一种带非均匀窗形式的金字塔状多层神经网络模型I研究了两种特征提取方法的向量规整问题,改进了Polak—Rjbiere学习算法并证明它能够保证连结权向量不收敛到非稳定局部极小点.特定人方式的试验表明,用神经网络方法识别五个元音和十个数字时,识别率都高于99.0%.用遍布25个省市的90个说话者的数字话音所进行的非特定人方式试验表明;加窗后识别率可以提高1.0%,达到90.3%,说明带预处理窗的模型更加适合于语音识别.

 
<< 更多相关文摘    
图标索引 相关查询

 


 
CNKI小工具
在英文学术搜索中查有关local minima的内容
在知识搜索中查有关local minima的内容
在数字搜索中查有关local minima的内容
在概念知识元中查有关local minima的内容
在学术趋势中查有关local minima的内容
 
 

CNKI主页设CNKI翻译助手为主页 | 收藏CNKI翻译助手 | 广告服务 | 英文学术搜索
版权图标  2008 CNKI-中国知网
京ICP证040431号 互联网出版许可证 新出网证(京)字008号
北京市公安局海淀分局 备案号:110 1081725
版权图标 2008中国知网(cnki) 中国学术期刊(光盘版)电子杂志社