基于谱分解的模糊C均值算法在彩色图像分割中的应用

针对模糊C均值聚类算法对初始值敏感、易陷入局部最优以及谱聚类算法无法处理样本量过大的问题,提出了一种将模糊C均值聚类算法与谱聚类算法相结合的模糊谱聚类算法应用于彩色图像分割。大致分为三步,第一步对图像进行预处理,将颜色空间由RGB空间转换为Lab空间;第二步对特征空间进行冗余模糊C均值聚类算法得到冗余类;第三步由冗余类的隶属度矩阵和聚类中心矩阵得到冗余类的特征空间,并根据贴进度和传递闭包将该特征空间转换为冗余类的相似度矩阵进行谱聚类,完成冗余类的合并。实验结果表明,与模糊C均值聚类算法相比,模糊谱聚类算法对于初始值敏感问题、易陷入局部最优以及只能识别团状的蔟得到了很好的解决,从而使彩色图像分割结果更加合理。     

基于模拟退火的K调和均值聚类算法

K均值算法是最通用的划分聚类算法,然而它有高度依赖初始值和收敛于局部最小的缺点,K调和均值算法采用数据点与所有聚类中心的距离的调和平均替代了数据点与聚类中心的最小距离,解决了K均值算法对初值敏感的问题。这样虽然解决初始值敏感问题,局部最小收敛问题仍然存在。为了获得全局最优解,提出一种新的算法：基于模拟退火算法的K调和均值聚类。该算法将一种优秀的随机搜索算法——模拟退火算法引入K调和均值聚类,来解决局部最小收敛的问题,并将改进后的算法用于IRIS数据集的聚类分析,聚类结果与K均值算法结果对比,证明了改进算法的优越性。     

融合K-调和均值的混沌粒子群聚类算法

针对K-调和均值和混沌粒子群聚类算法的优缺点,提出了一种融合K-调和均值的混沌粒子群聚类算法。首先通过K-调和均值方法把粒子群分成若干个子群体,每个粒子根据其个体极值和所在子种群的全局极值来更新位置。其次,算法中引入变尺度混沌变异,抑制了早熟收敛,提高了计算精度。实验证明,该算法可以有效地避免算法陷入局部最优,在保证收敛速度的同时增强了算法的全局搜索能力,明显改善了聚类效果。     

融合K-调和均值和模拟退火粒子群的混合聚类算法

针对K-调和均值和模拟退火粒子群聚类算法的优缺点,提出了1种融合K-调和均值和模拟退火粒子群的混合聚类算法。首先通过K-调和均值方法将粒子群分成若干个子群,每个粒子根据其个体极值和所在子种群的全局极值来更新位置。同时引入模拟退火思想,抑制了早期收敛,提高了计算精度。本文使用Iris、Zoo、Wine和Image Segmentation,4个数据库,以F-measure为评价聚类效果的标准,对混合聚类算法进行了验证。研究发现,该混合聚类算法可以有效地避免陷入局部最优,在保证收敛速度的同时增强了算法的全局搜索能力,明显改善了聚类效果。该算法目前已用于无锡一淡水养殖基地的水产健康养殖水质分析系统,运行效果良好。     

融合改进差分进化思想的K-调和均值聚类

针对K-调和均值聚类算法易陷入局部最优的缺点,提出了一种基于改进差分进化的K-调和均值聚类算法。该算法通过引入基于Logistic变尺度混沌搜索和指数递增交叉概率算子的差分进化算法来增强全局寻优能力。实验结果表明,该算法能够较好地克服K-调和均值算法的缺点,在保证收敛速度的同时增强了算法的全局搜索能力。     

模糊c-均值算法改进及其对卫星遥感数据聚类的对比

提出的改进的模糊c-均值聚类方法采用基于标准协方差矩阵的Mahalanobis距离，即椭球体聚类方法，这种聚类算法更接近遥感数据散点图的实际情况，从而可以显著提高聚类效果。对北京卫星ASTER数据的聚类分析实验表明，改进的模糊c-均值聚类方法的聚类效果要优于K-均值聚类方法和常规的模糊c-均值聚类方法。     

一种对类别属性进行聚类的算法

当查询比较模糊,检索到的结果文档中表达了对查询的不同解释时,就要根据用户的相关反馈对返回结果进行聚类,本章首先介绍了一种著名的基于划分的聚类方法 K-均值算法。这种算法虽然效果显著,却无法处理类别属性的聚类任务。因此,本文基于层次分类方法,设计了一种针对类别属性分类的聚类算法,使其聚类后的返回结果具有高正确率的特点。     

具有相似性传播的K-均值欠定盲分离算法

对于稀疏信源的欠定盲分离问题,混合矩阵的估计是至关重要的。为了提高估计性能,提出一种组合的聚类分析算法。首先,利用短时傅里叶变换把时域中的观测信号转变成频域中的稀疏信号,并通过数据的归一化把稀疏信号在频域的线性聚类转变成致密聚类。然后,利用相似性传播AP聚类方法搜索每个观测数据的邻域自动形成数据族的数量和相对应的关键数据。最后,以AP聚类的结果作为K-均值算法的初始值,对每类(族)数据的聚类中心进一步修正。仿真结果表明,组合聚类法能有效地提高混合矩阵的估计精度。把AP聚类和K-均值算法相结合的另一个优势是,能够克服经典K-均值算法需要事先知道信源数量和对数据的初始划分非常敏感的缺陷。     

基于粒子群优化算法的模糊聚类分析

提出一种基于模糊C-均值算法和粒子群优化算法的混合聚类算法,该算法利用粒子群优化算法全局寻优的特点,有效地克服了模糊C-均值算法对初始值敏感、易陷入局部最优的缺点.实验表明,该算法具备良好的聚类效果.     

融合并行混沌萤火虫算法的K-调和均值聚类

针对K-调和均值算法易陷于局部最优的缺点,提出一种基于改进萤火虫算法(firefly algorithm, FA)的K-调和均值聚类算法。将基于FA的粗搜索与基于并行混沌优化FA的精细搜索相结合,其中精细搜索部分首先通过FA搜索到当前最优解及次优解,然后通过改进的logistic映射与并行混沌优化策略产生混沌序列在其附近直接搜索,以增强算法的寻优性能。最终,将这种改进的FA用于K-调和均值算法聚类中心的优化。实验结果表明:该算法不但对几种测试函数具有更高的搜索精度,而且对6种数据集的聚类结果均有一定的改善,有效地抑制了K-调和均值算法陷于局部最优的问题,提高了聚类准确性和稳定性。     

初始化独立的谱聚类算法

谱聚类作为一种新颖的聚类算法近年来受到模式识别领域的广泛关注。针对传统谱聚类算法对初始中心敏感的特点,通过引入对初值不敏感的k-调和平均算法,提出一种初始化独立的谱聚类算法。在人工数据和真实数据上的实验表明,相比于传统的k-means算法、FCM算法和EM算法,改进算法在稳定性和聚类性能上有了显著的提高。     

K-Harmonic means type clustering algorithm for mixed datasets

K-means type clustering algorithms for mixed data that consists of numeric and categorical attributes suffer from cluster center initialization problem. The final clustering results depend upon the initial cluster centers. Random cluster center initialization is a popular initialization technique. However, clustering results are not consistent with different cluster center initializations. K-Harmonic means clustering algorithm tries to overcome this problem for pure numeric data. In this paper, we extend the K-Harmonic means clustering algorithm for mixed datasets. We propose a definition for a cluster center and a distance measure. These cluster centers and the distance measure are used with the cost function of K-Harmonic means clustering algorithm in the proposed algorithm. Experiments were carried out with pure categorical datasets and mixed datasets. Results suggest that the proposed clustering algorithm is quite insensitive to the cluster center initialization problem. Comparative studies with other clustering algorithms show that the proposed algorithm produce better clustering results.     

Self-tuning clustering for high-dimensional data

Spectral clustering is an important component of clustering method, via tightly relying on the affinity matrix. However, conventional spectral clustering methods 1). equally treat each data point, so that easily affected by the outliers; 2). are sensitive to the initialization; 3). need to specify the number of cluster. To conquer these problems, we have proposed a novel spectral clustering algorithm, via employing an affinity matrix learning to learn an intrinsic affinity matrix, using the local PCA to resolve the intersections; and further taking advantage of a robust clustering that is insensitive to initialization to automatically generate clusters without an input of number of cluster. Experimental results on both artificial and real high-dimensional datasets have exhibited our proposed method outperforms the clustering methods under comparison in term of four clustering metrics.     

Data clustering based on hybrid K-harmonic means and modifier imperialist competitive algorithm

Clustering is a process for partitioning datasets. Clustering is one of the most commonly used techniques in data mining and is very useful for optimum solution. K-means is one of the simplest and the most popular methods that is based on square error criterion. This algorithm depends on initial states and is easily trapped and converges to local optima. Some recent researches show that K-means algorithm has been successfully applied to combinatorial optimization problems for clustering. K-harmonic means clustering solves the problem of initialization using a built-in boosting function, but it is suffering from running into local optima. In this article, we purpose a novel method that is based on combining two algorithms; K-harmonic means and modifier imperialist competitive algorithm. It is named ICAKHM. To carry out this experiment, four real datasets have been employed whose results indicate that ICAKHM. Four real datasets are employed to measure the proposed method include Iris, Wine, Glass and Contraceptive Method Choice with small, medium and large dimensions. The experimented results show that the new method (ICAKHM) carries out better results than the efficiency of KHM, PSOKHM, GSOKHM and ICAKM methods.     

An efficient hybrid data clustering method based on K-harmonic means and Particle Swarm Optimization

Clustering is the process of grouping data objects into set of disjoint classes called clusters so that objects within a class are highly similar with one another and dissimilar with the objects in other classes. K-means (KM) algorithm is one of the most popular clustering techniques because it is easy to implement and works fast in most situations. However, it is sensitive to initialization and is easily trapped in local optima. K-harmonic means (KHM) clustering solves the problem of initialization using a built-in boosting function, but it also easily runs into local optima. Particle Swarm Optimization (PSO) algorithm is a stochastic global optimization technique. A hybrid data clustering algorithm based on PSO and KHM (PSOKHM) is proposed in this research, which makes full use of the merits of both algorithms. The PSOKHM algorithm not only helps the KHM clustering escape from local optima but also overcomes the shortcoming of the slow convergence speed of the PSO algorithm. The performance of the PSOKHM algorithm is compared with those of the PSO and the KHM clustering on seven data sets. Experimental results indicate the superiority of the PSOKHM algorithm.     

一种基于改进KH与KHM聚类的混合数据聚类算法

为解决K-means聚类对初始聚类中心敏感和易陷入局部最优的问题,提出一种基于改进磷虾群算法与K-harmonic means的混合数据聚类算法.提出一种具有莱维飞行和交叉算子的磷虾群算法以改进磷虾群算法易陷入局部极值和搜索效率低的不足,即在每次标准磷虾群位置更新后加入新的位置更新方法进一步搜索以提高种群的搜索能力,同时交替使用莱维飞行与交叉算子对当前群体位置进行贪婪搜索以增强算法的全局搜索能力.20个标准测试函数的实验结果表明,改进算法不易陷入局部最优解,可在较少的迭代次数下有效地搜索到全局最优解的同时保证算法的稳定性.将改进的磷虾群算法与K调和均值聚类融合,即在每次迭代后用最优个体或经过K调和均值迭代一次后的新个体替换最差个体.5个UCI真实数据集的测试结果表明:融合后的聚类算法能够克服K-means对初始聚类中心敏感的不足且具有较强的全局收敛性.     

基于改进引力搜索的混合K-调和均值聚类算法研究

为了解决聚类算法容易陷入局部最优的问题,以及增强聚类算法的全局搜索能力,基于KHM算法以及改进的引力搜索算法,本文提出一种混合K-调和均值聚类算法（G-KHM）。G-KHM算法具有KHM算法收敛速度快的优点,但同时针对KHM算法容易陷入局部最优解的问题,在初始化后数据开始搜索聚类中心时采用了一种基于对象多样性及收敛性增强的引力搜索算法,该方法改进了引力搜索算法容易失去种群多样性的缺点,并同时具有引力搜索算法较强的全局搜索能力,可以使算法收敛到全局最优解。仿真结果表明,G-KHM算法能有效地避免陷入局部极值,具有较强的全局搜索能力以及稳定性,并且相比KHM算法、K-mean聚类算法、C均值聚类算法以及粒子群算法,在分类精度和运行时间上表现出了更好地效果。     

Maneuvering target oriented digital camouflage image generation method

Since the digital camouflage generation method based on fixed single background image cannot fully meet the hidden need of maneuvering targets. In this paper, a method of generating digital camouflage image based on several background images is proposed. Firstly, according to the active range of the maneuvering target, several background images are collected. Secondly, in order to resolve the problem that the clustering algorithm is sensitive to the initial clustering center and easy to fall into local optimum, K-harmonic means (KHM) clustering algorithm is introduced and initial clustering center is determined based on color histogram. Again, KHM clustering is used to extract color features from several background images, the first clustering is extracted from a single background image, and the second clustering is extracted with several color features sets extracted from the first clustering of background images as input. Finally, the regular hexagon is used as the basic unit of the digital camouflage image to construct the spot template, and the digital camouflage image is generated by the lowest horizontal line algorithm. Example verification and camouflage effect detection show that the digital camouflage image generated by this method can effectively realize the concealment of maneuvering targets and has good camouflage effect.