Adaptive spatial information-theoretic clustering for image segmentation

The incorporation of spatial context into clustering algorithms for image segmentation has recently received a significant amount of attention. Many modified clustering algorithms have been proposed and proven to be effective for image segmentation. In this paper, we propose a different framework for incorporating spatial information with the aim of achieving robust and accurate segmentation in case of mixed noise without using experimentally set parameters based on the original robust information clustering (RIC) algorithm, called adaptive spatial information-theoretic clustering (ASIC) algorithm. The proposed objective function has a new dissimilarity measure, and the weighting factor for neighborhood effect is fully adaptive to the image content. It enhances the smoothness towards piecewise-homogeneous segmentation and reduces the edge blurring effect. Furthermore, a unique characteristic of the new information segmentation algorithm is that it has the capabilities to eliminate outliers at different stages of the ASIC algorithm. These result in improved segmentation result by identifying and relabeling the outliers in a relatively stronger noisy environment. Comprehensive experiments and a new information-theoretic proof are carried out to illustrate that our new algorithm can consistently improve the segmentation result while effectively handles the edge blurring effect. The experimental results with both synthetic and real images demonstrate that the proposed method is effective and robust to mixed noise and the algorithm outperforms other popular spatial clustering variants.     

改进k均值算法在网络入侵检测中的应用研究

k均值聚类算法在入侵检测中已经得到了广泛的研究。该文在k均值算法基础上,提出了改进的k均值算法。将k均值算法和改进的k均值算法分别应用于入侵检测。试验结果表明,改进后的k均值算法能够避免k均值算法固有的缺点,并且有比较高的检测性能。     

二分K均值聚类算法优化及并行化研究

二分K均值聚类算法在二分聚类过程中的初始质心选取速度方面存在不足。为此,提出以极大距离点作为二分聚类初始质心的思想,提升算法的运行速度。研究如何在群集系统中进行快速聚类,根据二分K均值聚类算法的特性,采用数据并行的思想和均匀划分的策略,对算法进行并行化处理。实验结果表明,改进后的算法能获得比较理想的加速比和较高的使用效率。     

基于临近像素空间距离的模糊C均值聚类算法

针对传统图像分割算法对不同类型噪声敏感性缺陷的问题,基于临近像素空间距离的模糊C均值聚类算法即SFCM (fuzzy C means clustering algorithm based on the space distance of the nearest pixels)算法,采用核化的空间距离公式,将点到点之间的距离转化为点到空间的距离,很好的平衡了考察像素点临近像素点的灰度信息与位置信息间的关系,进一步克服了临近像素的位置差异对考察像素影响不同的缺点.通过在合成图像和自然图像上的大量实验并与几个传统算法进行对比,不仅表现出了很强的抗干扰能力,提高了聚类精度,并且很好的保留了原图像边缘等细节信息,体现出了较强的鲁棒性.     

聚类分析中的差异性度量方法研究

距离与差异性度量是聚类分析中的基本概念,是许多聚类算法的核心内容。在经典的聚类分析中,度量差异性的指标是距离的简单函数。该文针对混合属性数据集,提出两种距离定义,将差异性度量推广成为距离、类大小等因素的多元函数,使得原来只适用于数值属性或分类属性数据的聚类算法可用于混合属性数据。实验结果表明新的距离定义和差异性度量方法可提高聚类的质量。     

基于视频聚类的关键帧提取算法

关键帧可以有效减少视频索引的数据量,是分析和检索视频的关键。在提取关键帧过程中,为了解决传统聚类算法对初始参数敏感的问题,提出了一种改进的基于视频聚类的关键帧提取算法。首先,提取视频帧的特征,依据帧间相似度,对视频帧进行层次聚类,并得到初始聚类结果;接着使用K-means算法对初始聚类结果进行优化,最后提取聚类的中心作为视频的关键帧。实验结果表明该方法可以大幅提高关键帧的准确率和查全率,能较好地表达视频的主要内容。     

路径聚类：在Ｗeb站点中的知识发现

用户对Ｗeb站点的访问代表了用对Ｗeb站点上页面的访问兴越,这种兴越程序可以通过用户对Ｗeb站点上页面的浏览顺序表现出来,在对Ｗeb站点的记问日志进行事务识别后,可以根据群体用户对Ｗeb站点的访问顺序进行聚类,即路径聚类,那么最终每一个聚类集就反映出该聚类集中的全体用户具有相似的访问兴越,为了得到这种根据用户访问兴越而对用户集的划分,提出了Ｋ－paths路径聚类方法,在这种方法中,根据用户的访问兴越定义了新的相似性测量手段和聚类中心,实验的结果是成功的。     

利用多进制小波变换和模糊c-均值聚类进行纹理分割

提出了一种基于多进制小波变换的多纹理特征提取方法,通过对小波系数的标准差作为纹理测度以生成特征向量,利用模糊c-均值聚类算法进行纹理分割,获得了较好的实验结果。     

基于敏感度的个性化(α,l)-匿名方法

目前多数隐私保护匿名模型不能满足面向敏感属性值的个性化保护需求,也未考虑敏感属性值的分布情况,易受相似性攻击.为此,提出基于敏感度的个性化(α,l)-匿名模型,通过为敏感属性值设置敏感度,并定义等敏感度组的概念,对等价类中各等敏感度组设置不同的出现频率,满足匿名隐私保护的个性化需求.通过限制等价类中同一敏感度的敏感属性值出现的总频率,控制敏感属性值的分布,防止相似性攻击.提出一种基于聚类的个性化(α,l)-匿名算法,实现匿名化处理.实验结果表明,该算法能以与其他l-多样性匿名模型近似的信息损失量和时间代价,提供更好的隐私保护.     

Stability of a hierarchical clustering

Clustering algorithms have the annoying habit of finding clusters even when the data are generated randomly. Verifying that potential clusterings are real in some objective sense is receiving more attention as the number of new clustering algorithms and their applications grow. We consider one aspect of this question and study the stability of a hierarchical structure with a variation on a measure of stability proposed in the literature.^(1,2)Our measure of stability is appropriate for proximity matrices whose entries are on an ordinal scale. We randomly split the data set, cluster the two halves, and compare the two hierarchical clusterings with the clustering achieved on the entire data set. Two stability statistics, based on the Goodman-Kruskal rank correlation coefficient, are defined. The distributions of these statistics are estimated with Monte Carlo techniques for two clustering methods (single-link and complete-link) and under two conditions (randomly selected proximity matrices and proximity matrices with good hierarchical structure). The stability measures are applied to some real data sets.