摘要
为了发现移动对象的迁徙轨迹和经停地,提出结合经停地检测算法和单链接聚类算法的方法。通过青海湖鸟类的历史位置信息验证该方法的准确性和有效性,并与应用于本领域的其他方法进行分析比较,如DBSCAN聚类算法、减聚类及模糊聚类算法。结果显示提出的方法能够克服对比算法仅考虑迁徙数据空间位置信息的缺点,准确有效地挖掘出鸟类经停地和迁徙轨迹。
In order to discover the migratory trajectory and stopover, this paper proposed a method combining stopover detection algorithm and single-linkage cluster algorithm. It verified the accuracy and effectiveness of the above method by the history location information of the birds in Qinghai Lake, and other methods applied to the field was analyzed and compared, such as DBSCAN clustering algorithm, subtractive clustering and fuzzy clustering algorithm. The result shows that the proposed method can overcome the shortcomings of conventional approaches that only consider the space location information of the migratory da- ta, and discover the bird stopover and migratory trajectory accurately and effectively.
出处
《计算机应用研究》
CSCD
北大核心
2013年第12期3690-3693,3697,共5页
Application Research of Computers
关键词
聚类分析
经停地检测算法
单链接算法
迁徙轨迹
经停地发现
clustering analysis
stopover detection algorithm
single linkage algorithm
migratroy trajectory
stopover discovery
