Deep Convolutional Gaussian Processes

Kenneth Blomqvist; Samuel Kaski; Markus Heinonen

doi:10.1007/978-3-030-46147-8_35

Deep Convolutional Gaussian Processes

Kenneth Blomqvist, Samuel Kaski, Markus Heinonen^*

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

7 Citations (Scopus)

Abstract

We propose deep convolutional Gaussian processes, a deep Gaussian process architecture with convolutional structure. The model is a principled Bayesian framework for detecting hierarchical combinations of local features for image classification. We demonstrate greatly improved image classification performance compared to current convolutional Gaussian process approaches on the MNIST and CIFAR-10 datasets. In particular, we improve state-of-the-art CIFAR-10 accuracy by over 10% points.

Original language	English
Title of host publication	Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Proceedings
Editors	Ulf Brefeld, Elisa Fromont, Andreas Hotho, Arno Knobbe, Marloes Maathuis, Céline Robardet
Pages	582-597
Number of pages	16
DOIs	https://doi.org/10.1007/978-3-030-46147-8_35
Publication status	Published - 1 Jan 2020
MoE publication type	A4 Article in a conference publication
Event	European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases - Wurzburg, Germany Duration: 16 Sep 2019 → 20 Sep 2019 Conference number: 16

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11907 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases
Abbreviated title	ECML PKDD
Country/Territory	Germany
City	Wurzburg
Period	16/09/2019 → 20/09/2019

Keywords

Convolutions
Gaussian processes
Variational inference

Access to Document

10.1007/978-3-030-46147-8_35

https://arxiv.org/abs/1810.03052

Cite this

Blomqvist, K., Kaski, S., & Heinonen, M. (2020). Deep Convolutional Gaussian Processes. In U. Brefeld, E. Fromont, A. Hotho, A. Knobbe, M. Maathuis, & C. Robardet (Eds.), Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Proceedings (pp. 582-597). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11907 LNAI). https://doi.org/10.1007/978-3-030-46147-8_35

Blomqvist, Kenneth ; Kaski, Samuel ; Heinonen, Markus. / Deep Convolutional Gaussian Processes. Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Proceedings. editor / Ulf Brefeld ; Elisa Fromont ; Andreas Hotho ; Arno Knobbe ; Marloes Maathuis ; Céline Robardet. 2020. pp. 582-597 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{0dfe24b4aa71497ba2fe25ee0771209a,

title = "Deep Convolutional Gaussian Processes",

abstract = "We propose deep convolutional Gaussian processes, a deep Gaussian process architecture with convolutional structure. The model is a principled Bayesian framework for detecting hierarchical combinations of local features for image classification. We demonstrate greatly improved image classification performance compared to current convolutional Gaussian process approaches on the MNIST and CIFAR-10 datasets. In particular, we improve state-of-the-art CIFAR-10 accuracy by over 10% points.",

keywords = "Convolutions, Gaussian processes, Variational inference",

author = "Kenneth Blomqvist and Samuel Kaski and Markus Heinonen",

year = "2020",

month = jan,

day = "1",

doi = "10.1007/978-3-030-46147-8_35",

language = "English",

isbn = "9783030461461",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "582--597",

editor = "Ulf Brefeld and Elisa Fromont and Andreas Hotho and Arno Knobbe and Marloes Maathuis and C{\'e}line Robardet",

booktitle = "Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Proceedings",

note = "European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML PKDD ; Conference date: 16-09-2019 Through 20-09-2019",

}

Blomqvist, K, Kaski, S & Heinonen, M 2020, Deep Convolutional Gaussian Processes. in U Brefeld, E Fromont, A Hotho, A Knobbe, M Maathuis & C Robardet (eds), Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11907 LNAI, pp. 582-597, European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, Wurzburg, Germany, 16/09/2019. https://doi.org/10.1007/978-3-030-46147-8_35

Deep Convolutional Gaussian Processes. / Blomqvist, Kenneth; Kaski, Samuel ; Heinonen, Markus.

Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Proceedings. ed. / Ulf Brefeld; Elisa Fromont; Andreas Hotho; Arno Knobbe; Marloes Maathuis; Céline Robardet. 2020. p. 582-597 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11907 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Deep Convolutional Gaussian Processes

AU - Blomqvist, Kenneth

AU - Kaski, Samuel

AU - Heinonen, Markus

N1 - Conference code: 16

PY - 2020/1/1

Y1 - 2020/1/1

N2 - We propose deep convolutional Gaussian processes, a deep Gaussian process architecture with convolutional structure. The model is a principled Bayesian framework for detecting hierarchical combinations of local features for image classification. We demonstrate greatly improved image classification performance compared to current convolutional Gaussian process approaches on the MNIST and CIFAR-10 datasets. In particular, we improve state-of-the-art CIFAR-10 accuracy by over 10% points.

AB - We propose deep convolutional Gaussian processes, a deep Gaussian process architecture with convolutional structure. The model is a principled Bayesian framework for detecting hierarchical combinations of local features for image classification. We demonstrate greatly improved image classification performance compared to current convolutional Gaussian process approaches on the MNIST and CIFAR-10 datasets. In particular, we improve state-of-the-art CIFAR-10 accuracy by over 10% points.

KW - Convolutions

KW - Gaussian processes

KW - Variational inference

UR - http://www.scopus.com/inward/record.url?scp=85084819961&partnerID=8YFLogxK

UR - https://arxiv.org/abs/1810.03052

UR - https://github.com/kekeblom/DeepCGP

U2 - 10.1007/978-3-030-46147-8_35

DO - 10.1007/978-3-030-46147-8_35

M3 - Conference contribution

AN - SCOPUS:85084819961

SN - 9783030461461

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 582

EP - 597

BT - Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Proceedings

A2 - Brefeld, Ulf

A2 - Fromont, Elisa

A2 - Hotho, Andreas

A2 - Knobbe, Arno

A2 - Maathuis, Marloes

A2 - Robardet, Céline

T2 - European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases

Y2 - 16 September 2019 through 20 September 2019

ER -

Blomqvist K, Kaski S , Heinonen M. Deep Convolutional Gaussian Processes. In Brefeld U, Fromont E, Hotho A, Knobbe A, Maathuis M, Robardet C, editors, Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Proceedings. 2020. p. 582-597. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-46147-8_35

Deep Convolutional Gaussian Processes

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this