查询词典 algorithm

First, the character clustering algorithm regards the character as our features and then it can cluster these sentences of our target words to the correct group without any other resources. The average precision is 66.1%. Second, in the concept clustering algorithm of the aggregate computing, we use HowNet as the knowledge base of our feature words and obtain the concepts of these words. We cluster the sentences which have the same or similar concepts of the feature words into the same group. And then we can complement some lacks of the character clustering algorithm. The average precision is 72.3%. Third, regarding the concept clustering algorithm of the sememe distance, we use the sememe distance to compute its concept similarity. It improved similarity measure of the concept clustering algorithm of the aggregate computing. It achieves 81% average precision and gets better cluster quality.

词形分群演算法不受语料资源限制，能将词形相似且词义相近的词汇所属的句子分到同一群，经过人工验证，得到了66.1%的平均正确率；基於集合计算的概念分群演算法使用了知网做为撷取特徵词汇的知识库，透过知网取得词汇的概念，将具相同或相似概念的特徵词汇所属的句子分成同一群，补足词形分群演算法的不足，得到72.3%的平均正确率；基於义原距离的概念分群演算法则利用义原间的距离计算特徵概念的相似度，进一步改善了基於集合计算的概念分群演算法在相似度衡量的问题，得到81%的平均正确率，达到更好的分群效果。

Three methods to build soft-sensor model of Radial Basis Function neural network (K-means clustering algorithm, mixed Genetic Algorithm, mixed hierarchical Genetic Algorithm) is investigated and verified in simulation.In the condition of setting hidden layer's node numbers, K-means clustering algorithm and mixed Genetic Algorithm can satisfy online prediction of component content.

研究了三种建立RBF网络软测量模型方法(K-均值聚类算法、混合遗传算法、混合递阶遗传算法)，并进行仿真试验验证。K-均值聚类算法与混合遗传算法在预先设定隐含层节点数的情况下能满足组分含量在线估计。

Firstly, on the basis of defining the characteristics of project and reviewing the mathematical models for RCPSP, we constructed the mathematical model for RCPSP under the condition that there is non-fixed duration and splittable tasks in project; In order to solve above model, we emphatically discussed the basic principle of the swarm intelligence algorithm--the particle swarm optimization algorithm as well as improvement research for current algorithm, and then proposed a new improved particle swarm optimization algorithm and analyzed its application process in the resource-constrained project scheduling problem; Lastly, in order to confirm the validity of the model and the algorithm, we took an experiment based on 108 instances that produced by problem instance generator which was developed by our team.

首先，在对RCPSP项目特征进行界定以及对项目进度优化模型回顾的基础之上，构建了基于活动工期不固定、任务可拆分特征的RCPSP的数学模型；为了求解上述模型，本文详细阐述粒子群优化算法(Particle Swarm Optimization，PSO)的基本原理以及当前对其改进的研究成果，提出了一种新的改进PSO算法，并对其在RCPSP的应用过程进行了设计；最后，为了检验模型的有效性及算法的优越性，基于梯队内部开发的案例生成器生成的108个算例，进行了仿真试验。

Researches on kernel clustering algorithms. After combining K-means clustering algorithm and the theory of kernel-based learning algorithms, we propose a fast kernel K-means clustering method which is based on CPD kernel. The experiment results indicate that the clustering effect of the algorithm is better than that of K-means algorithm, the clustering speed of the algorithm is also fast than that of K-means algorithm.

核聚类算法的研究：探讨了K-均值聚类算法，通过将核学习理论与K-均值聚类算法结合，提出了一种基于CPD核函数的快速核K-均值聚类算法，并将该算法与基于Mercer核的核聚类算法进行了比较，实验结果显示，我们的方法不仅比K-均值聚类算法的聚类效果好，而且聚类速度快。

Finally we also discuss the algorithm of syllable division, correcting algorithm, multithread algorithm and algorithm dealing with the rules of vowel and consonant harmony and the final vowel change in the Uyghur language.

此外，本论文叙述了分音节算法，替换算法，多线程处理算法"语音的和谐"处理算法和"语音的同化"处理算法等关键性的算法。

The experiment result of algorithm teaching indicates that, Through training reading algorithm ability, refining the algorithmic rule ability, algorithmic language ability, applied algorithm ability, then these abilities form the resultant of forces to improve algorithm ability as a whole.

算法教学实验结果表明：通过培养阅读算法能力、提炼"算理"能力、算法语言能力、应用算法能力，使之形成合力，能促使算法能力的整体提高。

The study has focused on: 1 The research background and the current state of the power measurement were analyzed systematically, and then wavelet-based power measurement algorithm was certified feasible for astable signals; 2 We developed a new method of RMS and power measurement based on Dmeyer wavelet. After that three other wavelet measurement approaches were compared with Dmeyer wavelet-based measurement approach. Then the impact on measurement errors of amplitude frequency property and energy leakage of wavelet filter banks were deeply researched, and some measures were proposed to reduce it; 3 We developed a new strategy of power measurement algorithm based on lifting wavelet, which had calculation advantage compared with the first wavelet, so that it was more suitable for the realization of hardware; 4 The running system of wavelet-based power measurement algorithm based on DSP was proposed, and then it was realized on ICETEK-VC5416-A EVM board; 5 Existing problems and the further research direction of wavelet-based power sub-band measurement algorithm was discussed.

本文对基于小波变换的功率测量算法展开深入的分析研究，主要内容有：1）系统地分析了功率分解测量的研究背景、国内外的研究现状，并从能量守恒的角度验证了小波功率分频带测量算法在非平稳信号条件下的可行性；2）首次提出了Dmeyer小波变换有效值与功率的测量方法，进而通过对该方法与3种已经提出的小波测量方法的对比分析，深入研究了小波滤波器组的幅频特性和能量泄漏特性对测量误差的影响，并针对电网谐波的特点提出了减小影响的相关措施；3）首次提出了二代小波功率测量算法，对比一代小波功率分频带测量算法，它大大减少了算法实现的运算量及内存的需求量，从而更适合于硬件的实时实现；4）提出了小波功率分频带测量算法的DSP实现方案，并在ICETEK-VC5416-A EVM板上进行了实现；5）探讨了小波功率分频带测量算法目前仍存在的问题，并提出了进一步的研究方向。

Based on X-filtered LMS algorithm and -filtered LMS algorithm adaptive inverse control, we use a new variable step size LMS algorithm. Adding little computation, variable step size LMS algorithm can result in fast convergence speed and low residual error simultaneously. The adaptive feedback control can counteract the beginning error of the system. The adaptive disturbance canceler can best erase noises and disturbances.

在原来的X-滤波LMS算法自适应逆控制和-滤波LMS算法自适应逆控制方法的基础上，引入了新的变步长LMS算法，在计算量增加不多的前提下，能同时获得较快的收敛速度和较小的稳态误差；引入自适应的反馈补偿控制克服了被控系统的直流零频漂移，使控制系统在初开始工作阶段快速收敛；还引入自适应扰动消除器，它能最大限度的消除扰动。

The concept of genetic algorithm, its configuration, operators and existing problems are introduced in this paper. A new algorithm for clustering analysis is presented based on genetic algorithm. There are two characteristics in our approaches. Firstly, the algorithm is the general-purposed and our clustering analyzer can cluster large data set with mixed numeric and categorical attributes.

4介绍遗传算法的概念、算法结构、运算算子及主要存在问题；提出了一种基于遗传算法的聚类新方法MGA，该方法有两个主要优点：一是通用性强，可以对包含数值属性和符号属性的大数据集进行聚类；二是提高了数据挖掘的效率与质量。

Meanwhile, the experiments show that the multi-level swarm intelligence refinement algorithm is better than the multi-level tabu search refinement algorithm, especially in view of the probability that the approximate global minimum partition of the coarser graph may be the local minimum partition of the finer graph. Furthermore, the experiments show that the multi-level ant colony optimization refinement algorithm achieves more performance improvement than the multi-level particle swarm optimization refinement algorithm, using the gain of vertex more effectively.

实验数据反映出，针对最小图上的全局最优剖分可能是初始图的局部最优剖分、粗化图上的全局最优剖分可能是细化图的局部最优剖分的情况，基于群智能的多水平迁移优化算法相比基于禁忌搜索的多水平迁移优化算具备更强的逃离局部最优的能力；基于蚁群的多水平迁移优化算法相比基于微粒群的多水平迁移优化算法，对收益值的启发信息更为有效地利用，取得最佳的性能改进。

Fostering the concept of multi-culture firmly and mastering the development of advanced culture correctly are conductive to resisting terrorism, separatism of nationalities and ultraism of religion, which is of great academic theoretical, immediate and profound historical si...

牢固树立多元文化的理念和正确把握先进文化的发展方向，有助于在世界范围内自觉地抵制恐怖主义，民族分裂主义和宗教极端主义。不仅具有重大的学术理论意义，而且也具有重大的现实意义和深远的历史意义。

But if anyone walks in the night, he stumbles, because the light is not in him.

11：10 若在黑夜行走，就必碰跌，因为那光不在他里面。