Samplers and Their Applications: A Survey.
ACM Comput. Surv.(2019)
摘要
The notion of Lp sampling, and corresponding algorithms known as Lp samplers, has found a wide range of applications in the design of data stream algorithms and beyond. In this survey, we present some of the core algorithms to achieve this sampling distribution based on ideas from hashing, sampling, and sketching. We give results for the special cases of insertion-only inputs, lower bounds for the sampling problems, and ways to efficiently sample multiple elements. We describe a range of applications of Lp sampling, drawing on problems across the domain of computer science, from matrix and graph computations, as well as to geometric and vector streaming problems.
更多查看译文
关键词
Lp sampling, data stream algorithms
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要