聚点

Two-phase-learning is adopted. The strongpoint is that intuitionistic knowledge can be used to control the segmentation by using input-weight and distant-weight. Examples prove that segmentation using SOFM network is right and efficient. Surface fitting is an absolutely necessarily process in REGM. Surface fitting includes algebraic and parametric surface fitting. Algebraic surface fitting has not extra requirement for data distribute format and boundary shape.

第三种方法把人工智能用于反向工程的数据分区，因为数据分区的实质是相似特征的聚类，而神经网络可以实现自组织特征映射，在本论文中第一次把自组织特征映射神经网络用于多视图点云的数据分区，通过神经网络对数据点的坐标、法向量等几何特征进行聚类，并且运用了快速稳定的训练方法——两阶段训练算法，最重要的是，通过距离权值和输入权值可以把用户观察所得的直观知识加入聚类过程。

In the paper, a modified Similarity Measure Function, called Dsim, is proposed in order to solve the problem of non-contrasting behavior high dimensional space, and then a learning algorithm, called DCFCM, is advanced through the combination of Dsim and Approximated K-median and applied in intrusion detection which solves the problem of sharp border effecting problems and realizes the detection of abnormal incurrence.

本文将首先提出新的相似度函数Dsim，有效地解决了高维空间聚类选维和降维问题，实现了高效的聚类了；接着将Dsim与近似K-medians算法相结合，提出了新的模糊聚类算法----DCFCM，并将其用于入侵检测。解决了由尖锐边界、孤立点所带来的误报警和漏报警问题，实现了对异常行为的检测。

Secondly, the relevance feedback approach requires relevant documents to be similar to each other. That is, they should cluster. Ideally, the term distribution in all relevant documents will be similar to that in the documents marked by the users, while the term distribution in all nonrelevant documents will be different from those in relevant documents. Things will work well if all relevant documents are tightly clustered around a single prototype, or, at least, if there are different prototypes, if the relevant documents have significant vocabulary overlap, while similarities between relevant and nonrelevant documents are small.

第二，相关反馈方法需要相关文档之间相似，即它们可以聚类，理想地，所有相关文档 term 分布应该相似，相关档即用户标记的文档，同时不相关文档应该与相关文档的 term 分布不同，如果所有的相关文档都很紧凑地聚在一个单一的中心点，或至少有不同的中心点，如果相关文档有明显的词汇重复，同时相关与不相关文档相似性比较小，它才是可行的。

After analyzing the deficiencies of the traditional clustering algorithm DBSCAN (Density Based Spatial Clustering of Applications with Noise), a line oriented clustering method based on DBSCAN was proposed. The object clustered changed from the point to the line.

分析了密度聚类算法的局限性，在此基础上提出了一种基于密度的面向线段的聚类方法，将DBSCAN中聚类的对象由点转变为线段。

In the rotation diagram of polytrioxane obtained by γ-irradiation of trioxane there appear some new diffraction spots in addition to that for ordinary polyoxymethylene.

在辐照聚合制得的聚三聚甲醛的迴转图上，除出现普通聚甲醛的衍射点外，还存在着一些新的衍射斑点。

Molecular dynamics simulations of NPT (constant pressure and constant temperature) dynamics are carried out to obtain specific volume as a function of temperature for polyethylene , polypropylene , poly-isobutylene , polyoxymethylene , and polydimethylsiloxane .

用NPT分子动力学模拟来获得聚乙烯、聚丙烯、聚异丁烯、聚甲醛和聚二甲基硅氧烷在不同温度下的特征体积。体积玻璃化转变温度通过对模拟得到的V-T作图的曲线的拐点决定。

The method includes the following steps: firstly, a spherical neighborhood of present sample points is established by using the geometric spherical-modelling theory and all the sample points contained in the spherical neighborhood are adopted as candidate neighbor points, thus not only preserving the effectivity of the dimension reduction capability when data sets are sparse but also getting the advantages of low-sensitivity to isolated points and good stability of the preserved topological structure; then a data relevance matrix more matching semantics can be obtained by relevance measurement based on route clusters to update the candidate neighbor points in the spherical neighborhood and optimize the regular neighborhood space of the present sample points, thus improving the phenomenon that the dimension reduction of sample sets provided with folded curved faces is apt to suffer the integrated-structure distortion in case of heterogeneous data distribution.

首先利用几何开球原理建立当前样本点的球状邻域，将包含在球状邻域内的所有样本点作为候选近邻点，不但能够保持在数据集稀疏情况下的降维性能的有效性，而且具有对孤立点敏感性不高、保留拓扑结构稳定性好的优点。然后利用基于路径聚类的相关性度量得到更符合语义的数据相关性矩阵，用来对球状邻域内的候选近邻点进行更新，优化当前样本点的规则邻域空间，改善了当数据不均匀分布时在带有折叠弯曲面的样本集上降维容易出现整体结构扭曲的现象。

Propose a new agglomerative hierarchical clustering based method to eliminate outliers,with clustering tree to identify outliers.

提出一种基于凝聚层次聚类消除孤立点的新方法，借助聚类树识别孤立点。

Result 1 Magnetic nanoparticles, magnetic nanoparticles modified with antisense oligodeoxynucleotide of human telomerase reverse transcriptase induced HL-60 tumor cells to apoptosis, we could see typical morphologic change of apoptosis cells: karyopyknosis, chromation"s condensing and aggregation in nuclear, forming crescent-shaped or annulus structures to lean on edge of cell nucleus"s membrane and posing apoptosis body by Atomic Force Microscope, Fluorescence microscope, transmission electron Microscope 2 There was a significant difference compared with control group(p.01), inhibition ratio had significant positive correlation with medication dosage and time ;during 0.8-8μM dosage amplitude, inhibition ratio accrescenced by dosages increasing. However, the inhibition ratio would decrease when dosage over 8-80μM.

结果 1 磁性纳米粒子、修饰有端粒酶反义寡核苷酸的磁性纳米粒子诱导HL-60细胞发生凋亡，原子力显微镜、光学显微镜、荧光显微镜和透射电镜下均观察到HL-60细胞呈现典型的凋亡细胞的形态变化：细胞核固缩，核内染色质浓缩、凝聚、形成新月形或环状结构紧靠在细胞核膜边缘，并形成凋亡小体。2 磁性纳米粒子、修饰有端粒酶反义寡聚脱氧核苷酸的磁性纳米粒子对HL-60肿瘤细胞的生长和增殖有明显的抑制作用，与对照组相比有显著性差异(p＜0.01)，在剂量为0.8-8μmol／L范围内，抑制率随剂量的增加而增加，当剂量超过8μmol／L时，抑制率反而下降；3 磁性纳米粒子、修饰有端粒酶反义寡聚脱氧核苷酸的磁性纳米粒子可增强p53基因的表达活性，引起DNA降解损伤，反向调节细胞周期活动，促使细胞从G0期进入G1期，抑制肿瘤细胞的生长。4 修饰有端粒酶反义寡聚脱氧核苷酸的量子点能通过内吞作用进入HL-60肿瘤细胞的细胞核，可以在细胞内进行定位和促进HL-60肿瘤细胞的凋亡。

Traditional grid clustering methods fail to consider the affect of neighbored grids , and may result in unsmoothed clustering , wrong judgement of clustering boundary , etc .

传统网格聚类方法由于没有考虑到相邻网格内的数据点对考查网格的影响，存在不能平滑聚类以及聚类边界判断不清的情况。

The split between the two groups can hardly be papered over.

这两个团体间的分歧难以掩饰。

This approach not only encourages a greater number of responses, but minimizes the likelihood of stale groupthink.

这种做法不仅鼓励了更多的反应，而且减少跟风的可能性。