【摘要】 PAM是基于k-中心值聚类的一种算法,在处理数据集的聚类问题时,具有良好的准确性和伸缩性。但PAM算法在随机选取初始中心点时存在不足,而且在处理存在孤立点或噪声的数据时算法不是很健壮。本文针对这两点不足,使用了数据场的概念对PAM聚类算法进行了有益的改进,提高了算法的准确性和处理孤立点或噪声的能力,使其更适合于对数据集的处理,提高了挖掘结果的质量。更多还原

【Abstract】 PAM is a k-mediod algorithm. It is efficient and flexible to handle data sets. However, there are inaccurate and insufficient in random choosing initial data center in algorithm PAM, and PAM is not robust to treat data sets in- cluding outlier or noise. Based on data field conception, an improved PAM is designed to solve these problems. This improved algorithm is more efficient and improves quality of result of data mining.更多还原