A simple algorithm for finding frequent elements in streams and bags
A simple algorithm for finding frequent elements in streams and bags
复制标题
DOI：
10.1145/762471.762473
复制
发表时间：
2003-03-01
影响因子：
1.8
通讯作者：
Papadimitriou, CH
中科院分区：
文献类型：
--
作者：
Karp, RM;Shenker, S;Papadimitriou, CH
我们提出了一种简单，精确的算法，用于在多键中识别频率超过阈值theta的项目。该算法需要两个通过线性时间和空间1/theta。第一张通过是一种在线算法，概括了一种众所周知的算法来找到多数元素，用于识别最多1/theta项目的一组，其中包括所有具有大于theta的项目。
We present a simple, exact algorithm for identifying in a multiset the items with frequency more than a threshold theta. The algorithm requires two passes, linear time, and space 1/theta. The first pass is an on-line algorithm, generalizing a well-known algorithm for finding a majority element, for identifying a set of at most 1/theta items that includes, possibly among others, all items with frequency greater than theta.