局部最优

K（n，n）的局部最优可靠性

K_ ( n , n ) Graph is Locally Optimally Reliable

传统K均值算法对初始聚类中心敏感，聚类结果随不同的初始输入而波动，容易陷入局部最优值。

Traditional K-Means algorithm is sensitive to the initial centers and easy to get stuck at locally optimal value .

然而，由于它采用K均值的迭代方式，所以它对初始原型敏感且容易陷入局部最优。

However , it applies K-means paradigm , so it is sensitive to initialization and converges to local optimum easily .

该算法用抗体表示函数优化解的可能模式，通过构造克隆选择算子完成全局和局部最优解的搜索，利用B细胞网络保持多种抗体并存。

In this algorithm , antibodies represented the possible optimization solutions of functions . Cloning operator was used to search the local optima and the global optimum .

针对多Agent联盟数量是Agent个数指数倍的问题，给出了基于局部最优Agent联盟结构生成算法&OCS算法。

To solve the problem of the number of coalition structure increasing rapidly , OCS algorithm & formation of agent coalition structure based on local optimum is given .

论文提出了一种基于粒子群的多目标优化算法，该算法采用Pareto支配关系来更新粒子的个体最优值和局部最优值，用存储池保存搜索过程中发现的非支配解；

This article presents a Particle Swarm Optimization ( PSO ) algorithm for multiobjective optimization problems .

但BP算法过度依赖于初始权值的选择，收敛速度缓慢且容易陷入局部最优。

But BP algorithm depends on the choice of initial weight value excessively , restrain the speed slowly and apt to fall into the local optimization .

因此，对于复杂多峰优化问题，DC不能在单一种群中并行地维持多个全局或局部最优解。

Thus DC is not capable of finding and maintaining multiple global or local optima in parallel in a single population when used to optimize complicated multimodal problems .

使用遗传算法（GA）与模糊c均值聚类（FCM）算法相结合的方法来设计码书，有效地克服了FCM算法容易陷入局部最优且对初始值敏感的缺点。

Codebook of vector quantization is designed through fuzzy c-mean clustering and genetic algorithm . This algorithm overcomes the shortcomings of FCM effectively .

SISO系统参数辨识的局部最优信号

Locally Optimal Signal for Parameter Identification of SISO System

局部最优选择中的关键问题是获取各QoS属性的权重，本文提出了一种组合赋权法。

The key problem of local optimal selection is to get the weight of QoS properties , thus a new combination weighting approach is proposed .

然而，由于BP算法采用的是梯度下降法，预测权值是从某一确定的点开始，即局部最优搜索。

However , due to the BP algorithm uses is the gradient descent method , forecast from a sure right value of PM , namely the local optimal search .

TS算法是一种新兴的现代启发式寻优技术，适合于求解组合优化问题，并能以很大的概率跳出局部最优解。

Tabu Search ( TS ) is a new heuristic searching technique which is suitable for solving combinatorial optimization problem .

实验证明新算法有效解决了调和K均值算法中簇个数需事先给定及聚类算法容易陷入局部最优的问题。

The result of experiment indicate that the new algorithm efficiently resolves the problems of KHM algorithm that the count of clusters need decide prior and it well reach local optimum result .

这即克服了MeanShift算法容易陷入局部最优的缺陷，也将当前帧的观测值信息融入到了粒子滤波，实现了两者的优势互补。

The improved algorithm not only overcome the defect of mean shift when target was partially or completely occlusion , but also introduces current observation to particle filter , and realizes the complementary advantages of both .

但鉴于BP算法存在收敛慢以及收敛于局部最优解等缺点，采用遗传算法来优化神经网络的权值和阀值，规避BP算法的缺陷。

But due to the BP algorithm defects like the slow convergence and converging to local optimal solution , this module makes use of genetic algorithm to optimize neural network weights and thresholds .

为解决QoS路由算法容易陷入局部最优的问题，将模拟退火方法与路由计算结合起来，提出一种新的组合优化算法。

Taking into account the local optimization of some QoS routing algorithms , this paper introduced a new combinatorial optimization algorithm , which combines the simulated annealing and k-shortest path algorithm .

实验表明，改进后的SOM算法缩短了网络的训练时间，并且不容易收敛到局部最优。

Experiments show that the improved SOM can reduce the training time of network and is not easily converge to a local optimum .

比如初始聚类数K要事先指定，初始聚类中心选择存在随机性，算法容易生成局部最优解，受孤立点的影响很大等。

For example we must choose the initial clustering number . The choice of initial clustering centre has randomness . The algorithm receives locally optimal solution easily , the effect of isolated point is serious .

支持向量机建立在统计学理论的VC维理论和结构风险最小化原则上，避免了局部最优解，并克服了维数的灾难。

As built on structural risk minimization and VC dimension theory of Statistical Learning , SVM is not easy to be run into local optimum and conquers curse of Dimensionality handily .

实施了最优保留策略，改进交叉和变异操作，并结合模拟退火算法（SA）的Metropolis判别准则的复制策略，使寻优过程能够跳出局部最优解，从而形成了混合遗传算法。

The hybrid genetic algorithm is formed combined with the Metropolis discrimination criteria of simulated annealing algorithm ( SA ), which can jump from local optimization solution .

该方法能获得先快后慢的权重变化率，并能有效地避免传统PSO算法的早熟和局部最优的问题。

This method could make the weight changing rate from fast to slow , and avioded the " premature " and local optimum problems of traditional PSO algorithm effectively .

实验表明，它既较好地解决了局部最优问题，又可以利用FCM的优点来提高整体的收敛速度。

Experiments show that the proposed method can solve the locally optimum problem preferably , and improve the converge speed in virtue of the advantage of FCM algorithm .

在大量的科学研究和工程应用实践中，人们发现遗传算法（GeneticAlgorithm，GA）存在一些问题，主要有早熟收敛、容易陷入局部最优、局部搜索能力较弱、收敛速度慢等。

Currently , Genetic Algorithm ( GA ) has been applied in many practices . Some major drawbacks of GA such as premature , sticks to local optimum easily , low local search ability and slow speed of convergence , have been found .

基于Lemke的优化算法拟合效果较理想，避免了局部最优，且辨识时间少；

Fitting results by Lemke optimization algorithm are better , avoiding local optimum and costing less time ;

并通过建立p-median模型分析局部最优路径下成本较低的空间定位问题。

And through establishes under the p-median model analysis partially optimal choice the cost low space orientation question .

本文在Hopfield神经网络优化方法的基础上，根据模拟退火算法逃离局部最优解的原理，提出了一种神经网络优化计算的新方法。

According to the principle of escaping from local optimal solutions of simulated annealing algorithm , a new neural computing method for optimization is proposed based on Hopfield neural network method .

试验证明了此方法的有效性，解决了LBG算法局部最优的局限，获取更接近全局最优的码书。

The experimental results prove that the algorithm has better improved the expected distortion by overcoming the local optimality of LBG algorithm .

并提出了一种新的变异概率的计算方法，引入了呈指数变化的系数km，使变异概率随进化代数呈递减趋势变化，保证了个体的进化能力，有效防止算法陷入局部最优。

To ensure individual ability of evolution and effectively prevent the algorithm into local optimal solution , this article introduces a new coefficient km which changes exponentially and makes mutation rate reduce along with the evolving algebra .

为此，在本文中，提出了一个快速的J均值算法，通过子划分的最小类内方差或最大类间方差对局部最优解的搜索范围进行约束。

Thus , in this paper , a fast J-means algorithm is proposed , which uses the minimum intra-clusters or the maximum inter-cluster variance of subpartition to constraint the range of searching a local optimal solution .