基于改进的DBSCAN算法的土壤肥力变化的分析研究  

Analysis and Research of the Soil Fertility Status Based on Improved DBSCAN Algorithm

作　　者：郭万春[1] 蔡丽霞[2] 陈航[2] 陈桂芬[2] 

机构地区：[1]长春金融高等专科学校,长春130022 [2]吉林农业大学,长春130118

出　　处：《计算机科学》2013年第11A期412-414,共3页Computer Science

基　　金：吉林省世行项目(2012Z04)资助

摘　　要：通常基于密度的DBSCAN算法可以有效地处理任意形状的簇,但由于时空数据具有明显的差异性,该算法不能综合分析土壤肥力状况。针对这一问题,提出了一种基于改进的DBSCAN算法来对农安镇土壤肥力状况进行分析研究。首先利用层次分析法得到土壤养分各属性的权值,以平衡数据间的差异性;其次,利用改进的DBSCAN算法对农安镇的土壤肥力数据进行分析,并将实验结果与传统的DBSCAN算法进行比较。实验结果表明,改进的DBSCAN算法对于选取Eps和minPts两个参数更加快速、有效,聚类结果更好。Typically DBSCAN algorithm based on density can effectively deal with clusters of arbitrary shape, but the algorithm is not a comprehensive analysis of soil fertility, Since temporal data have obvious differences. To solve this problem, this paper proposes a method to analyze the situation of soil fertility in nong＇an town based on improved DBSCAN algorithm. First, using AHP to get the weight of each property in order to balance the differences between the data;Secondly, the improved DBSCAN algorithm apply to analyze the data of soil fertility, and the experimental results with traditional DBSCAN algorithm were compared. Experimental results show that the improved DBSCAN algorithm more quickly and efficiently for selecting the two parameters Eps and minPts, it has better clustering results.

关 键 词：层次分析 DBSCAN算法 土壤肥力 权重值 

分 类 号：TP399[自动化与计算机技术—计算机应用技术] S158.2[自动化与计算机技术—计算机科学与技术]