引力势能聚类算法

Gravitational Potential Energy Clusting Algorithm

总结密度聚类算法存在的共性问题,即聚类之前的参数设定困难,据此提出密度聚类算法的改进目标。模拟万有引力势能的物理模型,结合核密度估计的概念,构建引力势能影响函数与引力势能密度函数,从而创造引力势能聚类算法,该算法能够克服聚类算法中的参数设定困难。详细介绍了该算法的基本原理、参数设定、聚类评判依据,算法步骤,并通过实际应用案例展示该算法在聚类分析和异常分析中的作用。

Summarized the common problem of the density-based clustering algorithm,which is the parameter settings before clustering,proposed to the improvement goals of density clustering algorithm.By simulating the physical model of Gravitational Potential Energy(GPE),combined with the concept of Kernel Density Estimate,created GPE Affect Function and GPE Density Function,thereby build the GPE Clustering Algorithm(GPECA),GPECA is able to overcome the problem of parameter set tings.Described theory,parameter settings,cluster determinations and the steps of GPECA.Demonstrated its applications in clus ter analysis and anomaly analysis through practical application cases.