一种基于CLIQUE聚类的全球地震目录分析方法

CLIQUE是一种重要的数据挖掘算法,广泛应用于大型数据库中的高维数据聚类。分析了CLIQUE算法的主要思想以及聚类算法在地震目录分析中的研究现状,提出了利用CLIQUE算法对全球地震目录进行聚类处理的流程。根据时空数据的多维特征,首先划分子空间计算密集单元,再将其连接聚簇并投影至各个维度进行可视分析。以近40 a(1977-2016年)的全球地震目录为数据源进行CLIQUE聚类实验,结果表明CLIQUE能有效发现地震现象在不同维度下呈现的聚集模式,且相对于其他聚类算法具有较高的效率。本文方法具有一定的可靠性与实用性,能够为地震事件的评估和防范提供决策依据。     

基于密度的加权K-Means算法

K-Means算法是比较流行的局域聚类算法,但由于其存在需要输入聚类数目以及对初始聚类中心敏感等缺陷,本文提出了一种基于密度的加权K-Means聚类算法来初始化聚类中心。该算法定义了点的密度函数和聚类中心函数,通过一定评价函数获取聚类中心。该方法获取的聚类中心不仅周围密度比较大,而且各个聚类中心之间相关性比较小,从而有效的减少了聚类时间,提高算法效率。     

数据质量聚类算法

在聚类算法中，聚类中心决定聚类的最终结果，而传统的分割聚类算法不能准确定位聚类中心。根据数据场提出了数据质量聚类中心的新概念，给出数据质量聚类算法，能够一次定位聚类中心，无需迭代，也无需预置聚类个数。7组对比实验表明，提出的方法能够准确定位聚类中心，获得良好的聚类结果和稳定性，优于传统的分割聚类算法和峰值密度聚类算法。     

基于聚类有效性函数的面状地理实体聚类

为解决聚类数未知条件下面状地理实体的聚类问题,文中提出了一种基于聚类有效性函数的聚类方法.给出了适合面状地理实体k-中心点聚类算法的聚类有效性函数;将该有效性函数改写为适应度函数,设计了基于遗传算法的面状地理实体聚类算法.该算法在计算聚类数的同时能得到划分聚类结果.实验结果从一定程度上反映了数据集的结构信息特征.     

基于聚类有效性函数的面状地理实体聚类

为解决聚类数未知条件下面状地理实体的聚类问题,文中提出了一种基于聚类有效性函数的聚类方法。给出了适合面状地理实体k-中心点聚类算法的聚类有效性函数;将该有效性函数改写为适应度函数,设计了基于遗传算法的面状地理实体聚类算法。该算法在计算聚类数的同时能得到划分聚类结果。实验结果从一定程度上反映了数据集的结构信息特征。     

一种基于属性值变化程度定权的聚类算法

针对经典K-means聚类算法以欧氏距离作为相似度判断法则进行聚类划分,而未考虑聚类对象的各属性值对聚类划分的影响程度存在差异的问题,该文提出了一种基于属性值变化程度定权的聚类算法。通过采用Iris dataset数据进行实验,该算法相对于其他聚类算法获得了更好的聚类效果,且该算法适用于生物物种分类、遥感影像识别等工作领域,能提高聚类运算的精准度。     

中心均匀化密度峰值聚类的激光点云分割

针对常规的密度峰值聚类算法在确定数据聚类中存在聚类中心的重复性、聚类不稳定、不适用于三维点云分割等问题,提出了中心均匀化聚类群融合算法.该算法对局部密度和距离函数进行归一化处理,较好地解决了这两种函数尺度不一的问题;基于局部密度和距离函数乘积的变化率来确定聚类中心,并对重复或距离很近的聚类中心进行了消除,避免了聚类中心非均匀分布对聚类的影响;利用数据点到聚类中心距离逐个确定每个数据的聚类归属,依据邻近聚类数据群之间的距离来判断邻近聚类之间的融合,实现对点云数据的有效分割.基于二维离散数据聚类及不同分辨率点云数据分割的实验结果表明:所提算法不仅适用于二维离散数据的聚类,也适用于三维点云数据的分割,且分割精度和稳定度要优于常规的CFDP、K-means、DBSCAN、DPC聚类算法和深度学习方法.     

时空大数据分布式增量IMSTDCA聚类方法研究

时空聚类分析是对时空大数据进行利用的一种有效手段,目前传统聚类算法存在着大规模分布数据难以处理,海量数据处理时间较长,确定参数困难,聚类质量较差等缺陷。因此,提出一种分布式增量聚类流程DICP,利用广域网分布增量聚类方法,避免大量数据的传输拷贝,有效提升聚类运算效率。对于DICP流程中的时空数据聚类算法本身,研究了一种大数据环境下的IMSTDCA时空数据聚类算法,借助密度聚类的思想,通过时空数据的聚集趋势预分析、时空数据聚类算法,以及时空数据聚类结果评价3个步骤完成聚类分析,实现时空大数据的快速高效信息挖掘。     

面向多维时空位置数据的动态加权聚类模型

针对传统聚类算法在处理时空位置数据挖掘时面临的多维聚类问题,提出了动态加权聚类模型。该模型叠加利用经典k-均值和基于密度的DBSCAN聚类算法,通过计算最大轮廓系数确定合适的簇数目,按照划分初始簇类、识别和剔除噪声点、修正聚类簇中心点位置坐标3个步骤实现对大体量多维时空位置数据的聚类分析,提出了动态权重系数计算公式,优化了基于密度的DBSCAN聚类算法中相似度函数,并在Python3.7环境下以网络签到数据集实例仿真验算了该模型算法。实验结果表明,相较单一的传统聚类算法,该模型能综合利用多维非位置属性对时空位置数据点聚类,更合理界定聚类簇的归属数据点,对提升时空位置数据集聚类簇中数据点的聚类效果明显。     

超谱遥感图像快速聚类无损压缩算法

K-means聚类要求每个像素要和所有聚类中心求欧氏距离，当聚类数很多时，这是一个相当耗时的工作。改进的K—meam聚类算法根据历史聚类结果进行初始类分割，即节约初始聚类时间，又能使历史聚类过程中形成的类间稳定关系得以保持；类内像素只和相邻的聚类中心计算距离进行聚类，随着算法的迭代进行，大量类的状态基本固定，使得聚类速度不断加快。基于改进K-means聚类的无损压缩算法具有充分利用历史聚类成果和收敛速度快的特点，通过提高类内像素冗余度，最大限度消除谱间冗余和空间冗余。采用多次聚类压缩的结果预测最佳聚类数的方法，可实现最小熵无损压缩。通过和DPCM算法概率模型的熵值比较及实验数据的分析，验证了基于聚类无损压缩效率比不聚类无损压缩效果更优。     

Combining Geo‐SOM and Hierarchical Clustering to Explore Geospatial Data

Geo‐SOM is a useful geovisualization technique for revealing patterns in spatial data, but is ineffective in supporting interactive exploration of patterns hidden in different Geo‐SOM sizes. Based on the divide and group principle in geovisualization, the article proposes a new methodology that combines Geo‐SOM and hierarchical clustering to tackle this problem. Geo‐SOM was used to “divide” the dataset into several homogeneous subsets; hierarchical clustering was then used to “group” neighboring homogeneous subsets for pattern exploration in different levels of granularity, thus permitting exploration of patterns at multiple scales. An artificial dataset was used for validating the method's effectiveness. As a case study, the rush hour motorcycle flow data in Taipei City, Taiwan were analyzed. Compared with the best result generated solely by Geo‐SOM, the proposed method performed better in capturing the homogeneous zones in the artificial dataset. For the case study, the proposed method discovered six clusters with unique data and spatial patterns at different levels of granularity, while the original Geo‐SOM only identified two. Among the four hierarchical clustering methods, Ward's clustering performed the best in pattern discovery. The results demonstrated the effectiveness of the approach in visually and interactively exploring data and spatial patterns in geospatial data.     

A comparison of MODIS, NCEP, and TMI sea surface temperature datasets

The monthly average sea surface temperature (SST) datasets of MODIS (Moderate Resolution Imaging Spectroradiometer), NCEP (National Center for Environmental Prediction) and TMI (Tropical Rainfall Measuring Mission (TRMM) Microwave Imager) are compared for the period March 2000 to June 2003. Large discrepancies (0.5 K->1 K) are found over extensive areas: the tropical Atlantic, tropical western Pacific, Bay of Bengal, Arabian Sea and the storm tracks. Many of these discrepancies are related to the biases inherent in the infrared and microwave retrieval methods. Probable causes for these biases include cirrus contamination, insufficient corrections for water vapor absorption and aerosol attenuation in infrared retrieval as well as uncertainty in surface emissivity in microwave retrieval. The SST difference patterns bear close resemblance to the patterns of distribution of aerosols, cirrus, atmospheric water vapor and surface wind speed at certain regions. Correlations between SST difference and aerosol optical depth, column water vapor and surface wind speed in some areas are high (>0.75). These biases have to be adjusted in order for the SST datasets to be more useful for climate studies.     

空间层次聚类显著性判别的重排检验方法

同时顾及空间邻近与专题属性相似的空间层次聚类是挖掘空间分布模式的一种有效手段。空间层次聚类方法虽然可以获得多层次的聚集结构,但聚类结果显著性的统计判别依然是一个尚未解决的难题。为此,本文提出了一种空间层次聚类结果显著性的统计判别方法,用于确定空间层次聚类的停止准则,减少聚类过程对参数设置的依赖。通过试验分析与比较发现,该方法能够有效判别空间层次聚类结果的显著性和确定层次聚类合并过程的停止条件,同时具有很好的抗噪性,避免随机结构的干扰。     

一种基于重排检验的时空聚类方法

融合时空邻近与专题属性相似的时空聚类是挖掘地理现象时空演化规律的重要手段。现有方法需要的聚类参数许多难以获取,影响了聚类方法的可操作性与聚类结果的可靠性。提出一种基于重排检验的时空聚类方法。首先,通过重排检验发现时空数据集中的均质子区域;进而,采用均方误差准则合并均质子区域内的时空实体生成时空簇,并通过簇内重排检验自动识别聚类合并的终止条件;最后,借助时空拓扑关系在保证结果精度的前提下发展一种快速重排检验的方法,提高了聚类方法的运行效率。通过实验和比较发现,该方法一方面可以发现不同形状、大小的时空簇,聚类质量优于经典的ST-DBSCAN方法;另一方面聚类过程中人为设置参数的主观性显著降低,提高了聚类方法的可操作性。     

尺度驱动的空间聚类理论

空间聚类是探索性空间数据分析的有力手段,不仅可以直接用于发现地理现象的分布格局与分布特征,亦可以为其他空间数据分析任务提供重要的预处理步骤。空间聚类有望成为大数据认知的突破口。空间聚类研究虽然已经引起了广泛关注,但是依然面临两大最根本的困境:"无中生有"和"无从理解"。"无中生有"指的是:绝大多数方法,即使针对不包含聚类结构的数据集,仍然会发现聚类;"无从理解"指的是:即使同一种聚类方法,采用不同的聚类参数就会获得千变万化的聚类结果,而这些结果的含义不明确。造成上述困境的根本原因在于:尺度没有在聚类模型中被当作重要参数而恰当地体现。为此,笔者受到人类视觉多尺度认知原理的启发,根据多尺度表达的"自然法则",建立了一套尺度驱动的空间聚类理论。首先将尺度定量化建模为聚类模型的参数,然后将空间聚类的尺度依赖性建模为一种假设检验问题,最后通过控制尺度参数以自动获得统计显著的多尺度聚类结果。在该理论指导下,可以构建适用不同应用需求的多尺度空间聚类模型,一方面降低了空间聚类过程中的主观性,另一方面有利于对空间聚类模式进行全面而深入的分析。     

Mining knowledge from result comparison between spatial clustering themes

This paper introduces some definitions and defines a set of calculating indexes to facilitate the research, and then presents an algorithm to complete the spatial clustering result comparison between different clustering themes. The research shows that some valuable spatial correlation patterns can be further found from the clustering result comparison with multi-themes, based on traditional spatial clustering as the first step. Those patterns can tell us what relations those themes have, and thus will help us have a deeper understanding of the studied spatial entities. An example is also given to demonstrate the principle and process of the method.     

Detection of Mesoscale Eddy-Related Structures Through Iso-SST Patterns

This letter, addressed to the analysis of remote sensing (RS) images of the sea-surface temperature (SST) off the Portugal coast, presents a novel approach to automatically detect and characterize mesoscale eddy-related structures. The complexity of this task is due to the dynamics of the investigated region, where upwelling currents and bathymetry effects produce countless and highly heterogeneous SST patterns, features of interest may have smooth boundaries, and edges associated to strong temperature gradients may not correspond to any eddy. All these limit the effectiveness of an image processing based on edge features (which can be successfully applied to automatically detect eddies in other oceanographic areas, for instance, close to the Gulf Stream). The proposed scheme exploits the iso-SST patterns associated to the eddy-related structure to code with a rule-based definition the process that allows for their visual identification. In practice, this enables revealing various morphological parameters of the eddy-related structure (i.e., the location, scale, symmetry, and rotation) and supports the exploitation of SST data allowing for annotating the RS image and benchmarking the subjectivity of the visual survey.     

移动轨迹聚类方法研究综述

轨迹数据是人类移动行为的表征,能够映射出人的出行模式和社会属性等信息。怎样有效挖掘轨迹数据蕴藏的人类活动规律一直是研究的热点。通过轨迹聚类发现行为相似的类簇,从而探究群体的移动模式是轨迹挖掘和深度应用常见的方法之一。本文首先根据轨迹数据的特点,将轨迹数据模型分为轨迹点模型和轨迹段模型,并据此定义相应的相似性度量：空间相似性度量和时空相似性度量;然后,对两类模型的聚类方法进行了综述,并总结不同聚类算法的优缺点,以期为不同应用选取聚类算法提供科学依据;最后对移动轨迹数据聚类方法研究的发展趋势进行了讨论。     

Trajectory analysis at intersections for traffic rule identification

In this paper,we focus on trajectories at intersections regulated by various regulation types such as traffic lights,priority/yield signs,and right-of-way rules.We test some methods to detect and recognize movement patterns from GPS trajectories,in terms of their geometrical and spatio-temporal components.In particular,we first find out the main paths that vehicles follow at such locations.We then investigate the way that vehicles follow these geometric paths(how do they move along them).For these scopes,machine learning methods are used and the performance of some known methods for trajectory similarity measurement(DTW,Hausdorff,and Fréchet distance)and clustering(Affinity propagation and Agglomerative clustering)are compared based on clustering accuracy.Afterward,the movement behavior observed at six different intersections is analyzed by identifying certain movement patterns in the speed-and time-profiles of trajectories.We show that depending on the regulation type,different movement patterns are observed at intersections.This finding can be useful for intersection categorization according to traffic regulations.The practicality of automatically identifying traffic rules from GPS tracks is the enrichment of modern maps with additional navigation-related information(traffic signs,traffic lights,etc.).     

On the spatial distribution of buildings for map generalization

Information on spatial distribution of buildings must be explored as part of the process of map generalization. A new approach is proposed in this article, which combines building classification and clustering to enable the detection of class differences within a pattern, as well as patterns within a class. To do this, an analysis of existing parameters describing building characteristics is performed via principal component analysis (PCA), and four major parameters (i.e. convex hull area, IPQ compactness, number of edges, and smallest minimum bounding rectangle orientation) are selected for further classification based on similarities between building characteristics. A building clustering method based on minimum spanning tree (MST) considering rivers and roads is then applied. Theory and experiments show that use of a relative neighbor graph (RNG) is more effective in detecting linear building patterns than either a nearest neighbor graph (NNG), an MST, or a Gabriel graph (GssG). Building classification and clustering are therefore conducted separately using experimental data extracted from OpenStreetMap (OSM), and linear patterns are then recognized within resultant clusters. Experimental results show that the approach proposed in this article is both reasonable and efficient for mining information on the spatial distribution of buildings for map generalization.