摘要
偏向参数和阻尼因子是影响AP聚类算法聚类效果的两个重要参数,但他们均取固定值。随着数据量的改变,原有参数取值不能使算法聚类结果达到最优。鉴此,本文提出自适应AP聚类算法,当数据量发生改变时,自动调整并获取最优的偏向参数和阻尼因子,最终得到最优聚类结果。与原来算法相比,改进后的算法能自动消除震荡,还可获取最优聚类结果,提高聚类结果的准确性和算法快速性。通过人造数据集和Iris数据集实验,证明了自适应AP聚类算法的有效性。
Bias parameter and damping factor are two important parameters that affect the clustering effect of AP clustering algorithm,but they both take fixed values.As the amount of data changes,the original parameter values cannot make the algorithm clustering result optimal.In this paper,an adaptive AP clustering algorithm is proposed.When the amount of data changes,it automatically adjusts and obtains the optimal bias parameters and damping factors,and finally obtains the optimal clustering result.Compared with the original algorithm,the improved algorithm can automatically eliminate the vibration,and can also obtain the optimal clustering results,which improves the accuracy of the clustering results and the speed of the algorithm.Experiments on artificial datasets and Iris datasets demonstrate the effectiveness of the adaptive AP clustering algorithm.
作者
赖健琼
Lai Jianqiong(School of Intelligent Technology,Tianfu College of Swufe,Mianyang,Sichuan 621000,China)
出处
《计算机时代》
2022年第4期38-42,共5页
Computer Era
