CFSBC： Clustering in High-Dimensional Space Based on Closed Frequent Item Set

Clustering in high-dimensional space is an important domain in data mining. It is the process of discovering groups in a high-dimensional dataset, in such way, that the similarity between the elements of the same cluster is maximum and between different clusters is minimal. Many clustering algorithms are not applicable to high dimensional space for its sparseness and decline properties. Dimensionality reduction is an effective method to solve this problem. The paper proposes a novel clustering algorithm CFSBC based onclosed frequent hemsets derived from association rule mining. which can get the clustering attributes with high efficiency. The algorithm has several advantages. First, it deals effectively with the problem of dimensionality reduction. Second, it is applicable to different kinds of attributes, Third, it is suitable for very large data sets. Experiment shows that the proposed algorithm is effective and efficient     

The refinement algorithm consideration in text clustering scheme based on multilevel graph

To construct a high efficient text clustering algorithm, the multilevel graph model and the refinement algorithm used in the uncoarsening phase is discussed, The model is applied to text clustering. The performance of clustering algorithm has to be improved with the refinement algorithm application, The experiment result demonstrated that the muhilevel graph text clustering algorithm is available,     

Neuromuscular synaptogenesis: clustering of acetylcholine receptors revisited

Clustering of neurotransmitter receptors in the postsynaptic membrane is critical for efficient synaptic transmission. During neuromuscular synaptogenesis, clustering of acetylcholine receptors (AChRs) is an early sign of postsynaptic differentiation. Recent studies have revealed that the earliest AChR clusters can form in the muscle independent of motorneurons. Neurally released agrin, acting through the muscle-specific kinase MuSK and rapsyn, then causes further clustering and localization of clusters underneath the nerve terminal. AChRs themselves are required for agrin-induced clustering of several postsynaptic proteins, most notably rapsyn. Once formed, AChR clusters are stabilized by several tyrosine kinases and by components of the dystrophin/utrophin glycoprotein complex, some of which also direct postnatal synaptic maturation such as formation of postjunctional folds. This review summarizes these recent results about AChR clustering, which indicate that early clustering can occur in the absence of nerves, that AChRs play an active role in the clustering process and that partly different mechanisms direct formation versus stabilization of AChR clusters. Received 10 April 2002; received after revision 4 June 2002; accepted 10 June 2002     

基于自适应混合变异进化策略的神经模糊系统及应用研究

针对现存神经模糊系统中存在的问题,提出了基于自适应混合变异进化策略的神经模糊系统采用改进的最近邻域聚类算法对输入空间进行模糊聚类,确定模糊规则数以及模糊规则前件,这样做精简了模糊规则,不会因输入变量的增加而造成"维数灾难”;采用自适应混合进化策略确定模糊规则的后件,明显提高了算法的收敛速度和精度.将本文提出的基于自适应进化策略的神经模糊系统用于某炼油厂航煤干点的软测量建模,结果表明,该系统具有结构简单、建模精度高、泛化能力强等优点.     

一种新的混合聚类算法

聚类是数据挖掘的主要技术之一,是一种无导师监督的模式识别方式。聚类分析就是按照数据间的相似程度,依据特定的准则将数据划分成不同子类。K-平均算法是经典的聚类算法。蚂蚁聚类算法是近来涌现的新的聚类算法,它通过模拟蚁群的智能行为进行聚类分析,已经在数据挖掘中得到应用。通过分析蚂蚁聚类算法和K-平均算法两种不同聚类算法的基本思想,将两种算法结合得到混合聚类算法,仿真实验证明混合聚类算法的算法性能优于蚂蚁算法和K-平均算法。     

基于最邻近聚类和向量模糊c-均值的混沌预测

针对混沌时间序列难预测的问题,提出一种新的基于最邻近聚类和向量模糊c-均值(FCMV)聚类算法的模糊建模方法。其前提参数辨识分两步,首先用最近邻聚类法初始划分输入空间,得到规则数及初始聚类中心,再用FCMV把具有相同收敛向量的聚类中心归到同一个区域来优化前一步得到的聚类中心,得到前提参数;采用递推最小二乘算法辨识模型的结论参数。最后通过对Mackey-Glass混沌时间序列的建模和预测验证了该方法的有效性与实用性。     

一种混合遗传禁忌搜索的模糊聚类算法

传统基于目标函数法的模糊聚类算法是一种迭代的"爬山"算法,容易陷入局部最优解.提出了基于遗传算法与禁忌搜索结合的模糊聚类算法,综合运用遗传算法的多出发点和禁忌搜索的记忆性来改善聚类的效果,并通过迭代的遗传禁忌搜索算法产生最优聚类中心,实验中分别通过人工数据和标准数据测试验证了该算法的有效性.     

远程教育中基于粗糙集的聚类算法

针对目前远程教育中个性化教学水平较低的问题,提出了一种基于粗糙集的Web学习者聚类算法,并应用粗糙集的约简方法解决了学习者特征数据中的属性冗余问题,提高了聚类算法的效率,从而提高了远程教学网站的个性化教学水平.     

基于自适应聚类的文本潜在主题的自动发现

提出了一种自适应于不同题材文本自动确定其包含的潜在主题数K的方法.考虑到大多数文本的潜在主题分布符合段落密度特性,提出以段落为中心的研究策略,通过采用基于K均值的聚类算法联同自定义判别函数的聚类分析方法,实现了段落自适应聚类下的文本潜在主题的自动发现.实验结果表明,该方法在一定程度上能有效处理普遍存在的文风自由且主题表达灵活多样的各式文本.     

基于启发式群聚算法的机器人全局任务调度策略

提出了一种基于启发式群聚算法的机器人全局任务调度策略，在任务划分阶段采用一种启发式群聚算法对随机划分的各子任务按能否使总的运行时间缩短进行各子任务间的聚合，以使各子任务粒度和相互之间的通讯量达到一种优化状态，并在此基础上对群聚之后的各子群任务采用集中式动态调度策略，在程序运行过程中实现各处理器的负载平衡，提高各处理器的利用率，缩短总的运行时间．在一个由5个DSP（digital signal processor）处理器组成的同构型松耦合MIMD（multi instruction multidata）并行处理平台上，对平面四自由度连杆机器人在关节锁定下的运动控制任务采用上述先随机划分，再聚合，最后集中式调度的方法进行了并行实时仿真实验，取得了满意的并行性能指标．