SIGKDD
2000
14 years 7 days ago
2000
This paper presents a simple new algorithm that performs k-means clustering in one scan of a dataset, while using a bu er for points from the dataset of xed size. Experiments show...