In this paper, a cluster validity index called CDV index is presented. The CDV index is capable of providing a quality measurement for the goodness of a clustering result for a data set. The CDV index is composed of t...In this paper, a cluster validity index called CDV index is presented. The CDV index is capable of providing a quality measurement for the goodness of a clustering result for a data set. The CDV index is composed of three major factors, including a statistically calculated external diameter factor, a restorer factor to reduce the effect of data dimension, and a number of clusters related punishment factor. With the calculation of the product of the three factors under various number of clusters settings, the best clustering result for some number of clusters setting is able to be found by searching for the minimum value of CDV curve. In the empirical experiments presented in this research, K-Means clustering method is chosen for its simplicity and execution speed. For the presentation of the effectiveness and superiority of the CDV index in the experiments, several traditional cluster validity indexes were implemented as the control group of experiments, including DI, DBI, ADI, and the most effective PBM index in recent years. The data sets of the experiments are also carefully selected to justify the generalization of CDV index, including three real world data sets and three artificial data sets which are the simulation of real world data distribution. These data sets are all tested to present the superior features of CDV index.展开更多
An asymptotic method has been developed for investigation of kinetics of formation of compact objects with strong internal bonds. The method is based on the uncertainty relation for a coordinate and a momentum in spac...An asymptotic method has been developed for investigation of kinetics of formation of compact objects with strong internal bonds. The method is based on the uncertainty relation for a coordinate and a momentum in space of sizes of objects (clusters) with strongly pronounced collective quantum properties resulted from exchange interactions of various physical nature determined by spatial scales of the processes under consideration. The proposed phenomenological approach has been developed by analogy with the all-known ideas about coherent states of quantum mechanical oscillator systems for which a product of coordinate and momentum uncertainties (dispersions) accepts the value, which is minimally possible within uncertainty relations. With such an approach the leading processes are oscillations of components that make up objects, mainly: collective nucleon oscillations in a nucleus and phonon excitations in a mesostructure crystal lattice. This allows us to consider formation and growth of subatomic and mesoscopic objects in the context of a single formalism. The proposed models adequately describe characteristics of formation processes of nuclear matter clusters as well as mesoscopic crystals having covalent and quasi-covalent bonds between atoms.展开更多
为了同时对数据对象和特征进行聚类分析以提高聚类准确率,在模糊紧致性和分离性算法(fuzzy compactness and separation,FCS)基础上,提出一种结合类内紧致性和类间分离性的模糊联合聚类算法(fuzzy compactness and separation co-cluste...为了同时对数据对象和特征进行聚类分析以提高聚类准确率,在模糊紧致性和分离性算法(fuzzy compactness and separation,FCS)基础上,提出一种结合类内紧致性和类间分离性的模糊联合聚类算法(fuzzy compactness and separation co-clustering,FCSCC)。该算法在FCS的基础上增加了对特征维度的隶属度关系与熵最大化原理,能够在数据对象和特征2个维度上同时聚类。为验证该算法的有效性,另选择了3种算法在5个数据集上进行了对比实验,结果表明,FCSCC算法的聚类准确率高于其他3种算法。展开更多
文摘In this paper, a cluster validity index called CDV index is presented. The CDV index is capable of providing a quality measurement for the goodness of a clustering result for a data set. The CDV index is composed of three major factors, including a statistically calculated external diameter factor, a restorer factor to reduce the effect of data dimension, and a number of clusters related punishment factor. With the calculation of the product of the three factors under various number of clusters settings, the best clustering result for some number of clusters setting is able to be found by searching for the minimum value of CDV curve. In the empirical experiments presented in this research, K-Means clustering method is chosen for its simplicity and execution speed. For the presentation of the effectiveness and superiority of the CDV index in the experiments, several traditional cluster validity indexes were implemented as the control group of experiments, including DI, DBI, ADI, and the most effective PBM index in recent years. The data sets of the experiments are also carefully selected to justify the generalization of CDV index, including three real world data sets and three artificial data sets which are the simulation of real world data distribution. These data sets are all tested to present the superior features of CDV index.
文摘An asymptotic method has been developed for investigation of kinetics of formation of compact objects with strong internal bonds. The method is based on the uncertainty relation for a coordinate and a momentum in space of sizes of objects (clusters) with strongly pronounced collective quantum properties resulted from exchange interactions of various physical nature determined by spatial scales of the processes under consideration. The proposed phenomenological approach has been developed by analogy with the all-known ideas about coherent states of quantum mechanical oscillator systems for which a product of coordinate and momentum uncertainties (dispersions) accepts the value, which is minimally possible within uncertainty relations. With such an approach the leading processes are oscillations of components that make up objects, mainly: collective nucleon oscillations in a nucleus and phonon excitations in a mesostructure crystal lattice. This allows us to consider formation and growth of subatomic and mesoscopic objects in the context of a single formalism. The proposed models adequately describe characteristics of formation processes of nuclear matter clusters as well as mesoscopic crystals having covalent and quasi-covalent bonds between atoms.
文摘为了同时对数据对象和特征进行聚类分析以提高聚类准确率,在模糊紧致性和分离性算法(fuzzy compactness and separation,FCS)基础上,提出一种结合类内紧致性和类间分离性的模糊联合聚类算法(fuzzy compactness and separation co-clustering,FCSCC)。该算法在FCS的基础上增加了对特征维度的隶属度关系与熵最大化原理,能够在数据对象和特征2个维度上同时聚类。为验证该算法的有效性,另选择了3种算法在5个数据集上进行了对比实验,结果表明,FCSCC算法的聚类准确率高于其他3种算法。
