CoRE kernels

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The term "CoRE kernel" stands for correlation resemblance kernel. In many real-world applications (e.g., computer vision), the data are often high-dimensional, sparse, and non-binary. We propose two types of (nonlinear) CoRE kernels for non-binary sparse data and demonstrate the effectiveness of the new kernels through a classification experiment. CoRE kernels are simple with no tuning parameters. However, training nonlinear kernel SVM can be costly in time and memory and may not be always suitable for truly large-scale industrial applications (e.g., search). In order to make the proposed CoRE kernels more practical, we develop basic probabilistic hashing (approximate) algorithms which transform nonlinear kernels into linear kernels.

Original languageEnglish (US)
Title of host publicationUncertainty in Artificial Intelligence - Proceedings of the 30th Conference, UAI 2014
PublisherAUAI Press
Pages496-504
Number of pages9
ISBN (Electronic)9780974903910
StatePublished - Jan 1 2014
Event30th Conference on Uncertainty in Artificial Intelligence, UAI 2014 - Quebec City, Canada
Duration: Jul 23 2014Jul 27 2014

Other

Other30th Conference on Uncertainty in Artificial Intelligence, UAI 2014
CountryCanada
CityQuebec City
Period7/23/147/27/14

Fingerprint

Computer vision
Industrial applications
Tuning
Data storage equipment
Experiments

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence

Cite this

Li, P. (2014). CoRE kernels. In Uncertainty in Artificial Intelligence - Proceedings of the 30th Conference, UAI 2014 (pp. 496-504). AUAI Press.
Li, Ping. / CoRE kernels. Uncertainty in Artificial Intelligence - Proceedings of the 30th Conference, UAI 2014. AUAI Press, 2014. pp. 496-504
@inproceedings{8b240c774d514f09a67d676e8cab5931,
title = "CoRE kernels",
abstract = "The term {"}CoRE kernel{"} stands for correlation resemblance kernel. In many real-world applications (e.g., computer vision), the data are often high-dimensional, sparse, and non-binary. We propose two types of (nonlinear) CoRE kernels for non-binary sparse data and demonstrate the effectiveness of the new kernels through a classification experiment. CoRE kernels are simple with no tuning parameters. However, training nonlinear kernel SVM can be costly in time and memory and may not be always suitable for truly large-scale industrial applications (e.g., search). In order to make the proposed CoRE kernels more practical, we develop basic probabilistic hashing (approximate) algorithms which transform nonlinear kernels into linear kernels.",
author = "Ping Li",
year = "2014",
month = "1",
day = "1",
language = "English (US)",
pages = "496--504",
booktitle = "Uncertainty in Artificial Intelligence - Proceedings of the 30th Conference, UAI 2014",
publisher = "AUAI Press",

}

Li, P 2014, CoRE kernels. in Uncertainty in Artificial Intelligence - Proceedings of the 30th Conference, UAI 2014. AUAI Press, pp. 496-504, 30th Conference on Uncertainty in Artificial Intelligence, UAI 2014, Quebec City, Canada, 7/23/14.

CoRE kernels. / Li, Ping.

Uncertainty in Artificial Intelligence - Proceedings of the 30th Conference, UAI 2014. AUAI Press, 2014. p. 496-504.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - CoRE kernels

AU - Li, Ping

PY - 2014/1/1

Y1 - 2014/1/1

N2 - The term "CoRE kernel" stands for correlation resemblance kernel. In many real-world applications (e.g., computer vision), the data are often high-dimensional, sparse, and non-binary. We propose two types of (nonlinear) CoRE kernels for non-binary sparse data and demonstrate the effectiveness of the new kernels through a classification experiment. CoRE kernels are simple with no tuning parameters. However, training nonlinear kernel SVM can be costly in time and memory and may not be always suitable for truly large-scale industrial applications (e.g., search). In order to make the proposed CoRE kernels more practical, we develop basic probabilistic hashing (approximate) algorithms which transform nonlinear kernels into linear kernels.

AB - The term "CoRE kernel" stands for correlation resemblance kernel. In many real-world applications (e.g., computer vision), the data are often high-dimensional, sparse, and non-binary. We propose two types of (nonlinear) CoRE kernels for non-binary sparse data and demonstrate the effectiveness of the new kernels through a classification experiment. CoRE kernels are simple with no tuning parameters. However, training nonlinear kernel SVM can be costly in time and memory and may not be always suitable for truly large-scale industrial applications (e.g., search). In order to make the proposed CoRE kernels more practical, we develop basic probabilistic hashing (approximate) algorithms which transform nonlinear kernels into linear kernels.

UR - http://www.scopus.com/inward/record.url?scp=84923294292&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84923294292&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84923294292

SP - 496

EP - 504

BT - Uncertainty in Artificial Intelligence - Proceedings of the 30th Conference, UAI 2014

PB - AUAI Press

ER -

Li P. CoRE kernels. In Uncertainty in Artificial Intelligence - Proceedings of the 30th Conference, UAI 2014. AUAI Press. 2014. p. 496-504