Standard

Simultaneous Perturbation Stochastic Approximation for Few-Shot Learning. / Boiarov, Andrei; Granichin, Oleg; Granichina, Olga.

European Control Conference 2020, ECC 2020. Institute of Electrical and Electronics Engineers Inc., 2020. p. 350-355 9143831 (European Control Conference 2020, ECC 2020).

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

Harvard

Boiarov, A, Granichin, O & Granichina, O 2020, Simultaneous Perturbation Stochastic Approximation for Few-Shot Learning. in European Control Conference 2020, ECC 2020., 9143831, European Control Conference 2020, ECC 2020, Institute of Electrical and Electronics Engineers Inc., pp. 350-355, 19th European Control Conference, ECC 2020, Saint Petersburg, Russian Federation, 12/05/20.

APA

Boiarov, A., Granichin, O., & Granichina, O. (2020). Simultaneous Perturbation Stochastic Approximation for Few-Shot Learning. In European Control Conference 2020, ECC 2020 (pp. 350-355). [9143831] (European Control Conference 2020, ECC 2020). Institute of Electrical and Electronics Engineers Inc..

Vancouver

Boiarov A, Granichin O, Granichina O. Simultaneous Perturbation Stochastic Approximation for Few-Shot Learning. In European Control Conference 2020, ECC 2020. Institute of Electrical and Electronics Engineers Inc. 2020. p. 350-355. 9143831. (European Control Conference 2020, ECC 2020).

Author

Boiarov, Andrei ; Granichin, Oleg ; Granichina, Olga. / Simultaneous Perturbation Stochastic Approximation for Few-Shot Learning. European Control Conference 2020, ECC 2020. Institute of Electrical and Electronics Engineers Inc., 2020. pp. 350-355 (European Control Conference 2020, ECC 2020).

BibTeX

@inproceedings{85c3952f97614db09669e007e5201675,
title = "Simultaneous Perturbation Stochastic Approximation for Few-Shot Learning",
abstract = "Few-shot learning is an important research field of machine learning in which a classifier must be trained in such a way that it can adapt to new classes which are not included in the training set. However, only small amounts of examples of each class are available for training. This is one of the key problems with learning algorithms of this type which leads to the significant uncertainty. We attack this problem via randomized stochastic approximation. In this paper, we suggest to consider the new multi-task loss function and propose the SPSA-like few-shot learning approach based on the prototypical networks method. We provide a theoretical justification and an analysis of experiments for this approach. The results of experiments on the benchmark dataset demonstrate that the proposed method is superior to the original prototypical networks.",
author = "Andrei Boiarov and Oleg Granichin and Olga Granichina",
year = "2020",
month = may,
language = "English",
isbn = "9783907144015",
series = "European Control Conference 2020, ECC 2020",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "350--355",
booktitle = "European Control Conference 2020, ECC 2020",
address = "United States",
note = "19th European Control Conference, ECC 2020 ; Conference date: 12-05-2020 Through 15-05-2020",
url = "https://ecc20.eu/",

}

RIS

TY - GEN

T1 - Simultaneous Perturbation Stochastic Approximation for Few-Shot Learning

AU - Boiarov, Andrei

AU - Granichin, Oleg

AU - Granichina, Olga

PY - 2020/5

Y1 - 2020/5

N2 - Few-shot learning is an important research field of machine learning in which a classifier must be trained in such a way that it can adapt to new classes which are not included in the training set. However, only small amounts of examples of each class are available for training. This is one of the key problems with learning algorithms of this type which leads to the significant uncertainty. We attack this problem via randomized stochastic approximation. In this paper, we suggest to consider the new multi-task loss function and propose the SPSA-like few-shot learning approach based on the prototypical networks method. We provide a theoretical justification and an analysis of experiments for this approach. The results of experiments on the benchmark dataset demonstrate that the proposed method is superior to the original prototypical networks.

AB - Few-shot learning is an important research field of machine learning in which a classifier must be trained in such a way that it can adapt to new classes which are not included in the training set. However, only small amounts of examples of each class are available for training. This is one of the key problems with learning algorithms of this type which leads to the significant uncertainty. We attack this problem via randomized stochastic approximation. In this paper, we suggest to consider the new multi-task loss function and propose the SPSA-like few-shot learning approach based on the prototypical networks method. We provide a theoretical justification and an analysis of experiments for this approach. The results of experiments on the benchmark dataset demonstrate that the proposed method is superior to the original prototypical networks.

UR - http://www.scopus.com/inward/record.url?scp=85090126401&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85090126401

SN - 9783907144015

T3 - European Control Conference 2020, ECC 2020

SP - 350

EP - 355

BT - European Control Conference 2020, ECC 2020

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 19th European Control Conference, ECC 2020

Y2 - 12 May 2020 through 15 May 2020

ER -

ID: 62141876