一种选择性加权聚类融合算法

传统的聚类融合方法通过融合所有成员实现融合,无法彻底消除劣质聚类成员对融合质量的影响,而从聚类成员的选择和加权两方面进行聚类融合,即先采用两两融合技术代替融合所有聚类结果进行聚类成员选择,然后进行基于属性的聚类成员加权,在理论上具有更好优越性。通过对真实数据和模拟数据的实验发现,该算法能有效处理聚类成员的质量差异,比传统聚类融合能得到更好的聚类结果,具有较好可扩展性。     

一种改进的基于属性重要性的加权聚类集成

聚类成员的质量问题是影响聚类集成效果的关键因素之一。目前针对于聚类集成的一致性函数研究较多,针对聚类成员的质量研究较少。本文提出了一种改进的基于属性重要性的加权聚类集成,根据互信息来度量两个聚类成员共享统计信息。并以此设计权重,对初始聚类结果进行加权,进而生成结果。该算法实验结果表明,同对比的聚类集成算法相比,所提出算法的聚类效果有所提高。     

选择性聚类融合新方法研究

针对传统选择性聚类融合算法不能消除劣质聚类成员的干扰以及聚类准确性不高等问题,提出了一种新的选择性加权聚类融合算法。算法中提出了基于聚类有效性评价方法的参照成员选择方法和联合聚类质量以及差异度的选择策略,然后还提出了基于容错关系信息熵的属性重要性加权方法。新算法有效地克服了传统选择性聚类融合算法的缺点,消除了劣质聚类成员的干扰,提高了聚类的准确性。大量的对比实验结果表明了算法的有效,且性能显著提高。     

分组选择聚类融合算法

针对聚类融合算法可能出现信息失真等问题,提出一种新的聚类融合算法.该算法兼顾聚类质量与成员多样性,采用一种新的相似性度量,并依据度量结果先对聚类成员进行剪辑操作,再分组、选择,最后根据每个聚类成员对于各类别的贡献设计了一种新的加权函数.与其他方法相比,该方法具有较好的稳定性和精确性.     

基于互信息和分形维数相结合的选择性聚类融合算法研究

针对传统聚类融合算法不能消除劣质聚类成员的干扰,以及聚类准确性不高等问题,提出一种基于分形维数的选择性聚类融合算法.该算法实现增量式聚类,能够发现任意形状的聚类.通过基于互信息计算权值的选择策略,选取部分优质聚类成员,再利用加权共协矩阵实现融合,获得最终的聚类结果.实验证明,与传统聚类融合算法相比,该算法提高了聚类质量,具有较好的扩展性.     

选择性聚类融合研究进展

传统的聚类融合方法通常是将所有产生的聚类成员融合以获得最终的聚类结果。在监督学习中,选择分类融合方法会获得更好的结果,从选择分类融合中得到启示,在聚类融合中应用这种方法被定义为选择性聚类融合。对选择性聚类融合关键技术进行了综述,讨论了未来的研究方向。     

去噪处理下的复杂网络聚类优化模型研究

由于当前已有模型未能对加权复杂网络数据进行去噪处理,导致网络聚类结果和聚类质量不理想,聚类实际运行时间上升.提出一种基于空间关联性的加权复杂网络聚类模型,通过法向张量投票技术组建投影时的局部坐标系,获取网络编码局部几何结构,利用邻居节点间感知数据的空间关联性过滤噪声数据,同时使用坐标更新算法调整噪声点位置,实现复杂网络...     

一种基于图论的加权聚类融合算法

现有聚类融合算法对混合属性数据进行处理的效果不佳,主要是融合后的结果仍存在一定的分散性。为解决这个问题,提出了一种基于图论的加权聚类融合算法,通过对数据集聚类得到聚类成员后,利用所设计的融合函数对各个数据对象赋予权重,同时通过设置各个数据对间边的权重来确定数据之间的关系,得到带权最近邻图,再用图论的方法进行聚类。实验表明,该算法的聚类精度和稳定性优于其他聚类融合算法。     

动态加权模糊核聚类算法

为了克服噪声特征向量对聚类的影响,充分考虑各特征向量对聚类结果的贡献度的不同,运用mercer核将待聚类的数据映射到高维空间,提出了一种新的动态加权模糊核聚类算法.该算法运用动态加权,自动消弱噪声特征向量在分类中的作用,在对数据没有任何先验信息的情况下,不仅能够准确划分线性数据,而且能够做到非线性划分非团状数据.仿真和实际数据分类结果表明,数据中的噪声对分类结果影响较小,该算法具有很高的实用性.     

一种基于图论的加权聚类融合算法

现有聚类融合算法对混合属性数据进行处理的效果不佳,主要是融合后的结果仍存在一定的分散性。为解决这个问题,提出了一种基于图论的加权聚类融合算法,通过对数据集聚类得到聚类成员后,利用所设计的融合函数对各个数据对象赋予权重,同时通过设置各个数据对间边的权重来确定数据之间的关系,得到带权最近邻图,再用图论的方法进行聚类。实验表明,该算法的聚类精度和稳定性优于其他聚类融合算法。     

基于多粒度粗糙集的聚类融合方法

现有的聚类融合算法从聚类成员的角度出发,若使用全部聚类成员则融合结果受劣质成员影响,对聚类成员进行选择再进行融合则选择的策略存在主观性。为在一定程度上避免这两种局限性,可以从元素的角度出发,提出一种新的聚类融合方法。通过多粒度决策不一致粗糙集来选择一部分类别确定的元素,再利用这部分元素进行聚类融合生成新的划分;多粒度决策不一致粗糙集模型能够刻画多粒度决策过程中属性一致而决策不一致的现象,提出了一种基于多粒度决策不一致的粗糙集模型,并给出了一种聚类融合方法。具体做法是：首先在数据集上多次使用K-means聚类算法,生成论域上的多个粒结构;其次对所有粒结构两两之间求粒间包含度,建立包含度矩阵,对矩阵使用Otsu算法计算阈值,得出多组满足阈值条件的信息粒,求解多粒度决策不一致下近似和上近似;最后分别处理下近似与边界域中元素的类别,从而获得了一个经过融合的聚类划分。实验结果表明,该方法能够有效改善聚类的结果,具有较高的时间效率,且算法具有较好的鲁棒性。     

一种改进的自适应聚类集成选择方法

针对自适应聚类集成选择方法（Adaptive cluster ensemble selection,ACES）存在聚类集体稳定性判定方法不客观和聚类成员选择方法不够合理的问题,提出了一种改进的自适应聚类集成选择方法（Improved ACES,IACES）.IACES依据聚类集体的整体平均归一化互信息值判定聚类集体稳定性,若稳定则选择具有较高质量和适中差异性的聚类成员,否则选择质量较高的聚类成员.在多组基准数据集上的实验结果验证了IACES方法的有效性:1）IACES能够准确判定聚类集体的稳定性,而ACES会将某些不稳定的聚类集体误判为稳定;2）与其他聚类成员选择方法相比,根据IACES选择聚类成员进行集成在绝大部分情况下都获得了更佳的聚类结果,在所有数据集上都获得了更优的平均聚类结果.     

基于MapReduce的FCM聚类集成算法

针对传统的聚类集成算法难以高效地处理海量数据的聚类分析问题,提出一种基于MapReduce的并行FCM聚类集成算法。算法利用随机初始聚心来获取具有差异化的聚类成员,通过建立聚类成员簇间OVERLAP矩阵来寻找逻辑等价簇,最后利用投票法共享聚类成员中数据对象的分类情况得出最终的聚类结果。实验证明,该算法具有良好的精确度,加速比和扩展性,具有处理较大规模数据集的能力。     

基于k-means聚类的神经网络分类器集成方法研究

针对差异性是集成学习的必要条件,研究了基于k-means聚类技术提高神经网络分类器集成差异性的方法。通过训练集并使用神经网络分类器学习算法训练许多分类器模型,在验证集中利用每个分类器的分类结果作为聚类的数据对象;然后应用k-means聚类方法对这些数据聚类,在聚类结果的每个簇中选择一个分类器代表模型,以此构成集成学习的成员;最后应用投票方法实验研究了这种提高集成学习差异性方法的性能,并与常用的集成学习方法bagging、adaboost进行了比较。     

基于相对密度和熵的混合属性聚类融合算法

混合属性聚类是近年来的研究热点,对于混合属性数据的聚类算法要求处理好数值属性以及分类属性,而现存许多算法没有很好得平衡两种属性,以至于得不到令人满意的聚类结果.针对混合属性,在此提出一种基于交集的聚类融合算法,算法单独用基于相对密度的算法处理数值属性,基于信息熵的算法处理分类属性,然后通过基于交集的融合算法融合两个聚类成员,最终得到聚类结果.算法在UCI数据集Zoo上进行验证,与现存k-prototypes与EM算法进行了比较,在聚类的正确率上都优于k-prototypes与EM算法,还讨论了融合算法中交集元素比的取值对算法结果的影响.     

Cluster ensemble framework based on the group method of data handling

Cluster ensemble is a powerful method for improving both the robustness and the stability of unsupervised classification solutions. This paper introduced group method of data handling (GMDH) to cluster ensemble, and proposed a new cluster ensemble framework, which named cluster ensemble framework based on the group method of data handling (CE-GMDH). CE-GMDH consists of three components: an initial solution, a transfer function and an external criterion. Several CE-GMDH models can be built according to different types of transfer functions and external criteria. In this study, three novel models were proposed based on different transfer functions: least squares approach, cluster-based similarity partitioning algorithm and semidefinite programming. The performance of CE-GMDH was compared among different transfer functions, and with some state-of-the-art cluster ensemble algorithms and cluster ensemble frameworks on synthetic and real datasets. Experimental results demonstrate that CE-GMDH can improve the performance of cluster ensemble algorithms which used as the transfer functions through its unique modelling process. It also indicates that CE-GMDH achieves a better or comparable result than the other cluster ensemble algorithms and cluster ensemble frameworks.     

To improve the quality of cluster ensembles by selecting a subset of base clusters

Conventional clustering ensemble algorithms employ a set of primary results; each result includes a set of clusters which are emerged from data. Given a large number of available clusters, one is faced with the following questions: (a) can we obtain the same quality of results with a smaller number of clusters instead of full ensemble? (b) If so, which subset of clusters is more efficient to be used in the ensemble? In this paper, these two questions are going to be answered. We explore a clustering ensemble approach combined with a cluster stability criterion as well as a dataset simplicity criterion to discover the finest subset of base clusters for each kind of datasets. Also, a novel method is proposed in order to accumulate the selected clusters and to extract final partitioning. Although it is expected that by reducing the size of ensemble the performance decreases, our experimental results show that our selecting mechanism generally lead to superior results.     

Local Averaging of Ensembles of LVQ-Based Nearest Neighbor Classifiers

Ensemble learning is a well-established method for improving the generalization performance of learning machines. The idea is to combine a number of learning systems that have been trained in the same task. However, since all the members of the ensemble are operating at the same time, large amounts of memory and long execution times are needed, limiting its practical application. This paper presents a new method (called local averaging) in the context of nearest neighbor (NN) classifiers that generates a classifier from the ensemble with the same complexity as the individual members. Once a collection of prototypes is generated from different learning sessions using a Kohonen's LVQ algorithm, a single set of prototypes is computed by applying a cluster algorithm (such as K-means) to this collection. Local averaging can be viewed either as a technique to reduce the variance of the prototypes or as the result of averaging a series of particular bootstrap replicates. Experimental results using several classification problems confirm the utility of the method and show that local averaging can compute a single classifier that achieves a similar (or even better) accuracy than ensembles generated with voting.     

一种基于二部图谱划分的聚类集成方法

将二部图模型引入聚类集成问题中,使用二部图模型同时建模对象集和超边集,充分挖掘潜藏在对象之间的相似度信息和超边提供的属性信息.设计正则化谱聚类算法解决二部图划分问题,在低维嵌入空间运行K-means++算法划分对象集,获得最终的聚类结果.在多组基准数据集上进行实验,实验结果表明所提出方法不仅能获得优越的结果,而且具有较高的运行效率.     

Representative Consensus from Limited‐Size Ensembles

Characterizing the uncertainty and extracting reliable visual information from ensemble data have been persistent challenges in various disciplines, specifically in simulation sciences. Many ensemble analysis and visualization techniques take a probabilistic approach to this problem with the assumption that the ensemble size is large enough to extract reliable statistical or probabilistic summaries. However, many real‐life ensembles are rather limited in size, with only a handful of members, due to various restrictions such as storage, computational power, or sampling limitations. As a result, probabilistic inference is subject to imprecision and can potentially result in untrustworthy information in the presence of a limited sample‐size ensemble. In this case, a more reliable approach is to fuse the information present in an ensemble with a limited number of members with minimal assumptions. In this paper, we propose a technique to construct a representative consensus that is particularly suited for ensembles of a relatively small size. The proposed technique casts the problem as an ordering problem in which at each point in the domain, the ensemble members are ranked based on the local neighborhood. This local approach allows us to provide shape and irregularity sensitivity. The local order statistics will then be fused to construct a global consensus using a Bayesian approach to ensure spatial coherency of the local information. We demonstrate the utility of the proposed technique using a synthetic and two real‐life examples.