Random Matrix-Improved Kernels For Large Dimensional Spectral Clustering

Hafiz Tiomoko Ali; Abla Kammoun; Romain Couillet

doi:10.1109/SSP.2018.8450705

Random Matrix-Improved Kernels For Large Dimensional Spectral Clustering

Hafiz Tiomoko Ali, Abla Kammoun, Romain Couillet

Computer, Electrical and Mathematical Sciences and Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

5 Scopus citations

Abstract

Leveraging on recent random matrix advances in the performance analysis of kernel methods for classification and clustering, this article proposes a new family of kernel functions theoretically largely outperforming standard kernels in the context of asymptotically large and numerous datasets. These kernels are designed to discriminate statistical means and covariances across data classes at a theoretically minimal rate (with respect to data size). Applied to spectral clustering, we demonstrate the validity of our theoretical findings both on synthetic and real-world datasets (here, the popular MNIST database as well as EEG recordings on epileptic patients).

Original language	English (US)
Title of host publication	2018 IEEE Statistical Signal Processing Workshop (SSP)
Publisher	Institute of Electrical and Electronics Engineers (IEEE)
Pages	453-457
Number of pages	5
ISBN (Print)	9781538615713
DOIs	https://doi.org/10.1109/SSP.2018.8450705
State	Published - Sep 7 2018

Bibliographical note

KAUST Repository Item: Exported on 2020-10-01
Acknowledgements: The work of R. Couillet and H. Tiomoko Ali is supported by the ANR Project RMT4GRAPH (ANR-14-CE28-0006).

Access to Document

10.1109/SSP.2018.8450705

https://repository.kaust.edu.sa/bitstream/10754/655594/1/random_matrix_improved_kernels.pdf

Cite this

@inproceedings{10179b4ff76545c0b7ff8ec1def42193,

title = "Random Matrix-Improved Kernels For Large Dimensional Spectral Clustering",

abstract = "Leveraging on recent random matrix advances in the performance analysis of kernel methods for classification and clustering, this article proposes a new family of kernel functions theoretically largely outperforming standard kernels in the context of asymptotically large and numerous datasets. These kernels are designed to discriminate statistical means and covariances across data classes at a theoretically minimal rate (with respect to data size). Applied to spectral clustering, we demonstrate the validity of our theoretical findings both on synthetic and real-world datasets (here, the popular MNIST database as well as EEG recordings on epileptic patients).",

author = "Ali, {Hafiz Tiomoko} and Abla Kammoun and Romain Couillet",

note = "KAUST Repository Item: Exported on 2020-10-01 Acknowledgements: The work of R. Couillet and H. Tiomoko Ali is supported by the ANR Project RMT4GRAPH (ANR-14-CE28-0006).",

year = "2018",

month = sep,

day = "7",

doi = "10.1109/SSP.2018.8450705",

language = "English (US)",

isbn = "9781538615713",

pages = "453--457",

booktitle = "2018 IEEE Statistical Signal Processing Workshop (SSP)",

publisher = "Institute of Electrical and Electronics Engineers (IEEE)",

}

TY - GEN

T1 - Random Matrix-Improved Kernels For Large Dimensional Spectral Clustering

AU - Ali, Hafiz Tiomoko

AU - Kammoun, Abla

AU - Couillet, Romain

N1 - KAUST Repository Item: Exported on 2020-10-01 Acknowledgements: The work of R. Couillet and H. Tiomoko Ali is supported by the ANR Project RMT4GRAPH (ANR-14-CE28-0006).

PY - 2018/9/7

Y1 - 2018/9/7

N2 - Leveraging on recent random matrix advances in the performance analysis of kernel methods for classification and clustering, this article proposes a new family of kernel functions theoretically largely outperforming standard kernels in the context of asymptotically large and numerous datasets. These kernels are designed to discriminate statistical means and covariances across data classes at a theoretically minimal rate (with respect to data size). Applied to spectral clustering, we demonstrate the validity of our theoretical findings both on synthetic and real-world datasets (here, the popular MNIST database as well as EEG recordings on epileptic patients).

AB - Leveraging on recent random matrix advances in the performance analysis of kernel methods for classification and clustering, this article proposes a new family of kernel functions theoretically largely outperforming standard kernels in the context of asymptotically large and numerous datasets. These kernels are designed to discriminate statistical means and covariances across data classes at a theoretically minimal rate (with respect to data size). Applied to spectral clustering, we demonstrate the validity of our theoretical findings both on synthetic and real-world datasets (here, the popular MNIST database as well as EEG recordings on epileptic patients).

UR - http://hdl.handle.net/10754/655594

UR - https://ieeexplore.ieee.org/document/8450705

UR - http://www.scopus.com/inward/record.url?scp=85053843540&partnerID=8YFLogxK

U2 - 10.1109/SSP.2018.8450705

DO - 10.1109/SSP.2018.8450705

M3 - Conference contribution

SN - 9781538615713

SP - 453

EP - 457

BT - 2018 IEEE Statistical Signal Processing Workshop (SSP)

PB - Institute of Electrical and Electronics Engineers (IEEE)

ER -

Random Matrix-Improved Kernels For Large Dimensional Spectral Clustering

Abstract

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this