國立虎尾科技大學 |

Private Computations on Streaming Data.

紀錄類型:	書目-語言資料,手稿 : Monograph/item
正題名/作者:	Private Computations on Streaming Data./
作者:	Garbe, Kevin Matthew.
面頁冊數:	1 online resource (79 pages)
附註:	Source: Dissertations Abstracts International, Volume: 85-09, Section: B.
Contained By:	Dissertations Abstracts International85-09B.
標題:	Computer science. -
電子資源:	click for full text (PQDT)
ISBN:	9798381968279

Private Computations on Streaming Data.
Garbe, Kevin Matthew.

Private Computations on Streaming Data. - 1 online resource (79 pages)

Source: Dissertations Abstracts International, Volume: 85-09, Section: B.

Thesis (Ph.D.)--University of California, Los Angeles, 2024.

Includes bibliographical references

We present a framework for privacy-preserving streaming algorithms which combine the memory-efficiency of streaming algorithms with strong privacy guarantees. These algorithms enable some number of servers to compute aggregate statistics efficiently on large quantities of user data without learning the user's inputs. While there exists limited prior work that fits within our model, our work is the first to formally define a general framework, interpret existing methods within this general framework, and develop new tools broadly applicable to this model. To highlight our model, we designed and implemented a new privacy-preserving streaming algorithm to compute heavy hitters, which are the most frequent elements in a data stream. We provide a performance comparison between our system and Poplar, the only other private statistics algorithm which supports heavy hitters. We benchmarked ours and Poplar's systems and provided direct performance comparisons within the same hardware platform. Of note, Poplar requires linear space compared to our poly-logarithmic space, meaning our system is the first to compute heavy hitters within the privacy-preserving streaming model. A small memory footprint allows our algorithm (among other benefits) to run efficiently on very large input sizes without running out of memory or crashing.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2024

Mode of access: World Wide Web

ISBN: 9798381968279Subjects--Topical Terms:

573171
Computer science.
Subjects--Index Terms:

AlgorithmIndex Terms--Genre/Form:

554714
Electronic books.

Private Computations on Streaming Data.
LDR:02668ntm a22003857 4500 001 1151706
005 20241118085739.5
006 m o d
007 cr mn ---uuuuu
008 250605s2024 xx obm 000 0 eng d
020 $a 9798381968279
035 $a (MiAaPQ)AAI31140399
035 $a AAI31140399
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Garbe, Kevin Matthew. $3 1478523
245 1 0 $a Private Computations on Streaming Data.
264 0 $c 2024
300 $a 1 online resource (79 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 85-09, Section: B.
500 $a Advisor: Ostrovsky, Rafail.
502 $a Thesis (Ph.D.)--University of California, Los Angeles, 2024.
504 $a Includes bibliographical references
520 $a We present a framework for privacy-preserving streaming algorithms which combine the memory-efficiency of streaming algorithms with strong privacy guarantees. These algorithms enable some number of servers to compute aggregate statistics efficiently on large quantities of user data without learning the user's inputs. While there exists limited prior work that fits within our model, our work is the first to formally define a general framework, interpret existing methods within this general framework, and develop new tools broadly applicable to this model. To highlight our model, we designed and implemented a new privacy-preserving streaming algorithm to compute heavy hitters, which are the most frequent elements in a data stream. We provide a performance comparison between our system and Poplar, the only other private statistics algorithm which supports heavy hitters. We benchmarked ours and Poplar's systems and provided direct performance comparisons within the same hardware platform. Of note, Poplar requires linear space compared to our poly-logarithmic space, meaning our system is the first to compute heavy hitters within the privacy-preserving streaming model. A small memory footprint allows our algorithm (among other benefits) to run efficiently on very large input sizes without running out of memory or crashing.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2024
538 $a Mode of access: World Wide Web
650 4 $a Computer science. $3 573171
650 4 $a Computer engineering. $3 569006
650 4 $a Information technology. $3 559429
653 $a Algorithm
653 $a Heavy hitters
653 $a Privacy-preserving streaming model
653 $a Memory-efficiency
655 7 $a Electronic books. $2 local $3 554714
690 $a 0984
690 $a 0489
690 $a 0464
710 2 $a ProQuest Information and Learning Co. $3 1178819
710 2 $a University of California, Los Angeles. $b Computer Science 0201. $3 1182286
773 0 $t Dissertations Abstracts International $g 85-09B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31140399 $z click for full text (PQDT)