Deep learning with differential Gaussian process flows

Pashupati Hegde; Markus Heinonen; Harri Lähdesmäki; Samuel Kaski

Deep learning with differential Gaussian process flows

Pashupati Hegde, Markus Heinonen, Harri Lähdesmäki, Samuel Kaski

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

35 Lataukset (Pure)

Abstrakti

We propose a novel deep learning paradigm of differential flows that learn a stochastic differential equation transformations of inputs prior to a standard classification or regression function. The key property of differential Gaussian processes is the warping of inputs through infinitely deep, but infinitesimal, differential fields, that generalise discrete layers into a dynamical system. We demonstrate state-of-the-art results that exceed the performance of deep Gaussian processes and neural networks

Alkuperäiskieli	Englanti
Otsikko	The 22nd International Conference on Artificial Intelligence and Statistic
Sivut	1-15
Sivumäärä	16
Vuosikerta	89
Tila	Julkaistu - huhtikuuta 2019
OKM-julkaisutyyppi	A4 Artikkeli konferenssijulkaisuussa
Tapahtuma	International Conference on Artificial Intelligence and Statistics - Naha, Japani Kesto: 16 huhtikuuta 2019 → 18 huhtikuuta 2019 Konferenssinumero: 22

Julkaisusarja

Nimi	Proceedings of Machine Learning Research
Kustantaja	PMLR
Vuosikerta	89
ISSN (elektroninen)	2640-3498

Conference

Conference	International Conference on Artificial Intelligence and Statistics
Lyhennettä	AISTATS
Maa	Japani
Kaupunki	Naha
Ajanjakso	16/04/2019 → 18/04/2019

Pääsy asiakirjaan

hegde19aAccepted author manuscript, 8,83 MB

Sormenjälki Sukella tutkimusaiheisiin 'Deep learning with differential Gaussian process flows'. Ne muodostavat yhdessä ainutlaatuisen sormenjäljen.

Science-IT
Mikko Hakala (Manager)
Perustieteiden korkeakoulu
Laitteistot/tilat: Facility

Siteeraa tätä

@inproceedings{800994a5a0814cacbf6a2ca49a218cef,

title = "Deep learning with differential Gaussian process flows",

abstract = "We propose a novel deep learning paradigm of differential flows that learn a stochastic differential equation transformations of inputs prior to a standard classification or regression function. The key property of differential Gaussian processes is the warping of inputs through infinitely deep, but infinitesimal, differential fields, that generalise discrete layers into a dynamical system. We demonstrate state-of-the-art results that exceed the performance of deep Gaussian processes and neural networks",

keywords = "gaussian process, Bayesian methods",

author = "Pashupati Hegde and Markus Heinonen and Harri L{\"a}hdesm{\"a}ki and Samuel Kaski",

year = "2019",

month = apr,

language = "English",

volume = "89",

series = "Proceedings of Machine Learning Research",

publisher = "PMLR",

pages = "1--15",

booktitle = "The 22nd International Conference on Artificial Intelligence and Statistic",

note = "International Conference on Artificial Intelligence and Statistics, AISTATS ; Conference date: 16-04-2019 Through 18-04-2019",

}

Hegde, P , Heinonen, M , Lähdesmäki, H & Kaski, S 2019, Deep learning with differential Gaussian process flows. julkaisussa The 22nd International Conference on Artificial Intelligence and Statistic. Vuosikerta 89, Proceedings of Machine Learning Research, Vuosikerta 89, Sivut 1-15, International Conference on Artificial Intelligence and Statistics, Naha, Japani, 16/04/2019. <https://arxiv.org/abs/1810.04066>

Deep learning with differential Gaussian process flows. / Hegde, Pashupati ; Heinonen, Markus ; Lähdesmäki, Harri ; Kaski, Samuel.

The 22nd International Conference on Artificial Intelligence and Statistic. Vuosikerta 89 2019. s. 1-15 (Proceedings of Machine Learning Research; Vuosikerta 89).

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

TY - GEN

T1 - Deep learning with differential Gaussian process flows

AU - Hegde, Pashupati

AU - Heinonen, Markus

AU - Lähdesmäki, Harri

AU - Kaski, Samuel

N1 - Conference code: 22

PY - 2019/4

Y1 - 2019/4

N2 - We propose a novel deep learning paradigm of differential flows that learn a stochastic differential equation transformations of inputs prior to a standard classification or regression function. The key property of differential Gaussian processes is the warping of inputs through infinitely deep, but infinitesimal, differential fields, that generalise discrete layers into a dynamical system. We demonstrate state-of-the-art results that exceed the performance of deep Gaussian processes and neural networks

AB - We propose a novel deep learning paradigm of differential flows that learn a stochastic differential equation transformations of inputs prior to a standard classification or regression function. The key property of differential Gaussian processes is the warping of inputs through infinitely deep, but infinitesimal, differential fields, that generalise discrete layers into a dynamical system. We demonstrate state-of-the-art results that exceed the performance of deep Gaussian processes and neural networks

KW - gaussian process

KW - Bayesian methods

UR - https://arxiv.org/abs/1810.04066

M3 - Conference contribution

VL - 89

T3 - Proceedings of Machine Learning Research

SP - 1

EP - 15

BT - The 22nd International Conference on Artificial Intelligence and Statistic

T2 - International Conference on Artificial Intelligence and Statistics

Y2 - 16 April 2019 through 18 April 2019

ER -

Deep learning with differential Gaussian process flows

Abstrakti

Julkaisusarja

Conference

Pääsy asiakirjaan

Science-IT

Siteeraa tätä