分群

The participants were fifth grade students and they were divided into high score group, middle score group, and low score group according to their achievement in the learning area of science and technology. There were two learning projects in two periods of experimental instruction with different participants: the first project was concave and convex lens and the second project was the investigation of sound and create instrument.

研究对象为国小五年级学生，以学生的自然与生活科技学习成就，将学生分为高分群、中分群、低分群，进行不同研究对象的二阶段实验教学，其学习主题：第一阶段为「凹透镜与凸透镜」，第二阶段为「声音的探讨与乐器制作」。

Forexample, the group 2, 4 and 10 of BOX-PCR were good agreement with group 5, 13 and12 of numerical analysis, and the group 3, 4 and 8 of RAPD were equal to group 6, 12and 8 of numerical analysis, and the group 5, 6 and 8 were identical with group 6, 12 and8 of numerical analysis, which indicated the 4 methods used could reveal the biodiversityamong the tested free-living nitrogen fixation strains.

总的来说，本研究所采用的表型分群和遗传型分群的结果在总体趋势上是一致的，BOX-PCR的群2、4和10分别与数值分类中的第5、13和12表观群一致；RAPD的群3、4和8分别与数值分类的群6、12和8相对应，16S rRNA PCR-RFLP中的群5、6和8则数值分类的群6、12和10完全吻合，说明本实验所获得的研究结果可信。

For a Hall π-subgroup in π-separable group and a given π-Fong character, we show that there always exists a normal subgroup, such that some irreducible constituent of this π-Fong character which restricts on the join of these two subgroups is a π-Fong character associated to this normal subgroup.

对于π-可分群的一个Hall π-子群和一个给定的π-Fong特征标，我们证明了总存在一个正规子群使得这个π-Fong特征标在这两个子群的交上的限制的某个不可约成分是联系这个正规子群的一个π-Fong特征标。

In this scheme weak voltage area are determined by the network sensitivity index method; Considering the relation between low voltage level and voltage collapse, it is put forward that the voltage magnitude variation rate is taken as the observed information for the grouping of weak voltage buses and weak voltage grouping are given; Taking geographic factors into consideration, a fuzzy maximal tree is established to group the weak voltage buses. The method to measure the incomplete observability of buses and the flow chart of PMU configuration are given.

应用网损灵敏度法来判断电压崩溃临界点和电压薄弱区域；考虑低电压水平与电压崩溃的关系，提出将电压幅值变化率作为电压薄弱母线分群所观察的信息，并且给出了电压薄弱分群指标，兼顾地域因素，采用模糊最大树聚类法将电压薄弱母线分群；给出了母线非完全可观测程度的度量方法以及 PMU 的配置流程。

Prain, Trifolium repens Lam, Aeachynomene Mica L, Indigofera bungeana Steud.The same strains used in numerical taxonomy were analysed by BOXAIR-PCR fingerprinting. All tested strains were clustered into 47 groups at the similarity level of 50%. Compared with the result of numerical taxonomy, strains belonged to one phenon in numerical taxonomy always distributed into several BOXAIR-PCR groups, most of reference strains come from one recognized rhizobial species were not grouped together.

B(来源：dd9dAB94C论文网www.abclunwen.com)OXAIR-PCR指纹图谱的聚类结果显示：全部供试菌株在50％的相似性水平上分为47群，与数值分类的结果相比较，其分群的结果更分散，在数值分类80％相似性水平上位于同一表观群的供试未知菌株大多分布于几个综合的BOXAIR-PCR群内；另外在数值分类中来自同一个种的已知根瘤菌参比菌株几乎都聚在同一个表观群，而在BOXAIR-PCR分析中，已知参比根瘤菌的同一个种内的大部分菌株都单独成群。

Customer segmentation is a critical step to the success of target or segment marketing. The effectiveness of customer segmentation depends on a good number of variables such as customers' demographic, behavioral, geographic, and psychographic characteristics. However, the huge combinations and correlations among variables make customer segmentation a challenging task, even with the aids of information technologies. The industry is desperate for effectual and powerful tool for excavating customer knowledge from the large accumulated data in the corporate database management system.

摘要顾客分群对於目标行销或市场切割的成功与否都是一个重要的关键步骤，顾客分群结果的优劣取决於其所采用的分群因子，顾客分群中常使用的分群因子有顾客的人口统计资讯、行为特徵、地理性资讯与心理方面的特徵等，然而，这大量与因子复合交错的资讯使得顾客分群成为企业的一个挑战，并且渴望寻求一个能够从大量累积的资料库系统中挖掘知识的工具。

After that, I use those attributes to run cluster analysis with all the scattered sherds.

为了了解陶容器能否有效建构，笔者以社口遗址出土的陶质资料进行研究，先以经过修复组合成陶容器的样本为例，对其中的陶片进行属性登录，寻找最适合用来进行陶容器单位建构的分群属性组合；其次，将所选出的属性组合利用统计的集群分析方法套用在未知群组的陶片样本中，依照分群的结果建构出陶容器单位，并且根据样本分布的状况检视陶容器单位的建构成果是否能被认定为有效。

The study proposed three methods which are the character clustering algorithm, the concept clustering algorithm of the aggregate computing and the concept clustering algorithm of the sememe distance.

本研究提出三种分群方法，分别是词形分群演算法、基於集合计算的概念分群演算法和基於义原距离的概念分群演算法。

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。

However, when sitting for the learning achievement posttests, the experiment group, especially the mediocre and low achievers perform better than the control group. Therefore, we propose that the Embedded-Composite Assessment in the teaching is helpful for the experiment group to learn Factorisation, especially for the experiment group's mediocre and low achievers.

但在学习成就测验后测及学习成就测验前后测进步成绩上，实验组全组及中低分群都比对照组全组及中低分群的分数来得高，由此可推测「融入复式评量之实验教学」对实验组及实验组中低分群学习「因式分解」是有帮助的。

Neither the killing of Mr Zarqawi nor any breakthrough on the political front will stop the insurgency and the fratricidal murders in their tracks.

在对危险的南部地区访问时，他斥责什叶派民兵领导人对中央集权的挑衅行为。

In fact,I've got him on the satellite mobile right now.

实际上 我们已接通卫星可视电话了