2017
DOI: 10.1007/s13218-017-0519-3
|View full text |Cite
|
Sign up to set email alerts
|

Coresets-Methods and History: A Theoreticians Design Pattern for Approximation and Streaming Algorithms

Abstract: We present a technical survey on the state of the art approaches in data reduction and the coreset framework. These include geometric decompositions, gradient methods, random sampling, sketching and random projections. We further outline their importance for the design of streaming algorithms and give a brief overview on lower bounding techniques.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
43
0

Year Published

2019
2019
2022
2022

Publication Types

Select...
4
2
1

Relationship

1
6

Authors

Journals

citations
Cited by 50 publications
(43 citation statements)
references
References 75 publications
0
43
0
Order By: Relevance
“…Coreset construction techniques (Bachem et al, 2017 ; Munteanu and Schwiegelshohn, 2018 ) seek to create a “summary” weighted sample of a dataset with the property that a model learned on this dataset approximates one learned on the complete dataset. Here too, the difference in objectives is that we focus on small models, ignore training data size, and are interested in outperforming a model learned on the complete data.…”
Section: Overviewmentioning
confidence: 99%
“…Coreset construction techniques (Bachem et al, 2017 ; Munteanu and Schwiegelshohn, 2018 ) seek to create a “summary” weighted sample of a dataset with the property that a model learned on this dataset approximates one learned on the complete dataset. Here too, the difference in objectives is that we focus on small models, ignore training data size, and are interested in outperforming a model learned on the complete data.…”
Section: Overviewmentioning
confidence: 99%
“…), which turns out in fact to be a strong requirement. The reader interested in an overview of coreset construction techniques is referred to the recent review [99].…”
Section: Definitionmentioning
confidence: 99%
“…Meanwhile, research on data summarization has inspired a third approach: collecting data summaries. Data summaries, e.g., coresets, sketches, projections [18], [19], [20], are derived datasets that are much smaller than the original dataset, and can hence be transferred to a central location with a low communication overhead. This approach has been adopted in recent works, e.g., [6], [7], [8], [21].…”
Section: A Related Workmentioning
confidence: 99%
“…Because of the dependence on the cost function (Definition II.1), existing coreset construction algorithms are tailormade for specific machine learning problems. Here we briefly summarize common approaches for coreset construction and representative algorithms, and refer to [18], [19] for detailed surveys.…”
Section: B Coreset Construction Algorithmsmentioning
confidence: 99%
See 1 more Smart Citation