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^*

^*Tämän työn vastaava kirjoittaja

Tutkimustuotos: Artikkeli kirjassa/konferenssijulkaisussa › Conference contribution › Scientific › vertaisarvioitu

Abstrakti

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.

Alkuperäiskieli	Englanti
Otsikko	Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Proceedings
Toimittajat	Ulf Brefeld, Elisa Fromont, Andreas Hotho, Arno Knobbe, Marloes Maathuis, Céline Robardet
Kustantaja	SPRINGER
Sivut	582-597
Sivumäärä	16
ISBN (painettu)	9783030461461
DOI - pysyväislinkit	https://doi.org/10.1007/978-3-030-46147-8_35
Tila	Julkaistu - 1 tammikuuta 2020
OKM-julkaisutyyppi	A4 Artikkeli konferenssijulkaisuussa
Tapahtuma	European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases - Wurzburg, Saksa Kesto: 16 syyskuuta 2019 → 20 syyskuuta 2019

Julkaisusarja

Nimi	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Vuosikerta	11907 LNAI
ISSN (painettu)	0302-9743
ISSN (elektroninen)	1611-3349

Conference

Conference	European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases
Lyhennettä	ECML PKDD
Maa	Saksa
Kaupunki	Wurzburg
Ajanjakso	16/09/2019 → 20/09/2019

Pääsy asiakirjaan

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

Link to publication in Scopus

Sormenjälki Sukella tutkimusaiheisiin 'Deep Convolutional Gaussian Processes'. Ne muodostavat yhdessä ainutlaatuisen sormenjäljen.

Siteeraa tätä

Blomqvist, K., Kaski, S., & Heinonen, M. (2020). Deep Convolutional Gaussian Processes. teoksessa U. Brefeld, E. Fromont, A. Hotho, A. Knobbe, M. Maathuis, & C. Robardet (Toimittajat), Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Proceedings (Sivut 582-597). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vuosikerta 11907 LNAI). SPRINGER . 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. Toimittaja / Ulf Brefeld ; Elisa Fromont ; Andreas Hotho ; Arno Knobbe ; Marloes Maathuis ; Céline Robardet. SPRINGER , 2020. Sivut 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 = "1",

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)",

publisher = "SPRINGER",

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",

}

Blomqvist, K, Kaski, S & Heinonen, M 2020, Deep Convolutional Gaussian Processes. julkaisussa U Brefeld, E Fromont, A Hotho, A Knobbe, M Maathuis & C Robardet (toim), 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), Vuosikerta 11907 LNAI, SPRINGER , Sivut 582-597, European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, Wurzburg, Saksa, 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. toim. / Ulf Brefeld; Elisa Fromont; Andreas Hotho; Arno Knobbe; Marloes Maathuis; Céline Robardet. SPRINGER , 2020. s. 582-597 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vuosikerta 11907 LNAI).

Tutkimustuotos: Artikkeli kirjassa/konferenssijulkaisussa › Conference contribution › Scientific › vertaisarvioitu

TY - GEN

T1 - Deep Convolutional Gaussian Processes

AU - Blomqvist, Kenneth

AU - Kaski, Samuel

AU - Heinonen, Markus

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

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

PB - SPRINGER

ER -

Blomqvist K, Kaski S , Heinonen M. Deep Convolutional Gaussian Processes. julkaisussa Brefeld U, Fromont E, Hotho A, Knobbe A, Maathuis M, Robardet C, toimittajat, Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2019, Proceedings. SPRINGER . 2020. s. 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