2007
DOI: 10.1007/s00778-007-0065-y
|View full text |Cite
|
Sign up to set email alerts
|

Maintaining bounded-size sample synopses of evolving datasets

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

1
66
0

Year Published

2009
2009
2020
2020

Publication Types

Select...
3
3
2

Relationship

1
7

Authors

Journals

citations
Cited by 65 publications
(67 citation statements)
references
References 30 publications
1
66
0
Order By: Relevance
“…There has been recent work on maintaining bounded-size samples in a data warehouse that continually receives new data, but only random samples were considered [8,17]. Our (one-pass and two-pass) algorithms can be thought of as constructing special summaries that may be stored in a data warehouse for CFD confidence estimation.…”
Section: Related Workmentioning
confidence: 99%
“…There has been recent work on maintaining bounded-size samples in a data warehouse that continually receives new data, but only random samples were considered [8,17]. Our (one-pass and two-pass) algorithms can be thought of as constructing special summaries that may be stored in a data warehouse for CFD confidence estimation.…”
Section: Related Workmentioning
confidence: 99%
“…This means that reservoir sampling can handle only updates and insertions, and not the deletions. To deal with this issue, a new sampling method is proposed in [66], called Random Pairing (RP), to cope with deletions in a data set with stable size. In this way that, they add new data points to the sample to keep the size of sample constant when a deletion occurs in the sample.…”
Section: Random Pairingmentioning
confidence: 99%
“…There is more research considering maintenance of dynamic data streams such as [87]. Discovery of association rules through sampling has been investigated in many studies by [88][89][90] sampling approaches for database files are reviewed.…”
Section: Other Improvements and Applications Of Probability Sampling mentioning
confidence: 99%
“…"Priority Sampling" aims to reduce the variance on subset-sum queries [10]. There has also been work on sampling from streams which include deletions [15,16].…”
Section: · · ·mentioning
confidence: 99%