Streaming Quantiles Algorithms with Small Space and Update Time.

Autor: Ivkin N; Amazon, New York, NY 10001, USA., Liberty E; Pinecone, San Mateo, CA 94402, USA., Lang K; Yahoo Research, Sunnyvale, CA 94089, USA., Karnin Z; Amazon, New York, NY 10001, USA., Braverman V; Department of Computer Science, Rice University, Houston, TX 77005, USA.
Jazyk: angličtina
Zdroj: Sensors (Basel, Switzerland) [Sensors (Basel)] 2022 Dec 08; Vol. 22 (24). Date of Electronic Publication: 2022 Dec 08.
DOI: 10.3390/s22249612
Abstrakt: Approximating quantiles and distributions over streaming data has been studied for roughly two decades now. Recently, Karnin, Lang, and Liberty proposed the first asymptotically optimal algorithm for doing so. This manuscript complements their theoretical result by providing a practical variants of their algorithm with improved constants. For a given sketch size, our techniques provably reduce the upper bound on the sketch error by a factor of two. These improvements are verified experimentally. Our modified quantile sketch improves the latency as well by reducing the worst-case update time from O(1ε) down to O(log1ε).
Databáze: MEDLINE
Externí odkaz:
Nepřihlášeným uživatelům se plný text nezobrazuje