Future Generation Computer Systems, volume 127, pages 109-125

Automated evolutionary approach for the design of composite machine learning pipelines

Nikitin Nikolay O ¹

Barabanova Irina V ¹

Vychuzhanin Pavel ¹

Sarafanov Mikhail ¹

Polonskaia Iana S ¹

Maximov Gleb

Boukhanovsky A. ¹

Kalyuzhnaya Anna V

Revin Ilia ¹

Kalyuzhnaya Anna ¹

Hide authors affiliations Show authors affiliations: 1 affiliation

ITMO University, Saint-Petersburg, Russia |

Publication type: Journal Article

Publication date: 2022-02-01

Elsevier

Journal: Future Generation Computer Systems

Quartile SCImago

Quartile WOS

Impact factor: 7.5

ISSN: 0167739X

DOI: 10.1016/j.future.2021.08.022

Copy DOI

Hardware and Architecture

Computer Networks and Communications

Software

Abstract

The effectiveness of the machine learning methods for real-world tasks depends on the proper structure of the modeling pipeline. The proposed approach is aimed to automate the design of composite machine learning pipelines, which is equivalent to computation workflows that consist of models and data operations. The approach combines key ideas of both automated machine learning and workflow management systems. It designs the pipelines with a customizable graph-based structure, analyzes the obtained results, and reproduces them. The evolutionary approach is used for the flexible identification of pipeline structure. The additional algorithms for sensitivity analysis, atomization, and hyperparameter tuning are implemented to improve the effectiveness of the approach. Also, the software implementation on this approach is presented as an open-source framework. The set of experiments is conducted for the different datasets and tasks (classification, regression, time series forecasting). The obtained results confirm the correctness and effectiveness of the proposed approach in the comparison with the state-of-the-art competitors and baseline solutions.

By date By citations

Citations by journals

	1 2 3 4
Lecture Notes in Computer Science	Lecture Notes in Computer Science, 4, 14.81% Lecture Notes in Computer Science 4 publications, 14.81%
Procedia Computer Science	Procedia Computer Science, 2, 7.41% Procedia Computer Science 2 publications, 7.41%
Computers and Geosciences	Computers and Geosciences, 1, 3.7% Computers and Geosciences 1 publication, 3.7%
Water (Switzerland)	Water (Switzerland), 1, 3.7% Water (Switzerland) 1 publication, 3.7%
Systems Science and Control Engineering	Systems Science and Control Engineering, 1, 3.7% Systems Science and Control Engineering 1 publication, 3.7%
Journal of Personalized Medicine	Journal of Personalized Medicine, 1, 3.7% Journal of Personalized Medicine 1 publication, 3.7%
Knowledge-Based Systems	Knowledge-Based Systems, 1, 3.7% Knowledge-Based Systems 1 publication, 3.7%
Results in Engineering	Results in Engineering, 1, 3.7% Results in Engineering 1 publication, 3.7%
IFIP Advances in Information and Communication Technology	IFIP Advances in Information and Communication Technology, 1, 3.7% IFIP Advances in Information and Communication Technology 1 publication, 3.7%
Advances in Intelligent Systems and Computing	Advances in Intelligent Systems and Computing, 1, 3.7% Advances in Intelligent Systems and Computing 1 publication, 3.7%
ACS applied materials & interfaces	ACS applied materials & interfaces, 1, 3.7% ACS applied materials & interfaces 1 publication, 3.7%
AI Communications	AI Communications, 1, 3.7% AI Communications 1 publication, 3.7%
Annals of Operations Research	Annals of Operations Research, 1, 3.7% Annals of Operations Research 1 publication, 3.7%
Lecture Notes in Networks and Systems	Lecture Notes in Networks and Systems, 1, 3.7% Lecture Notes in Networks and Systems 1 publication, 3.7%
Journal of Information and Intelligence	Journal of Information and Intelligence, 1, 3.7% Journal of Information and Intelligence 1 publication, 3.7%
	1 2 3 4

Citations by publishers

	1 2 3 4 5 6 7 8
Springer Nature	Springer Nature, 8, 29.63% Springer Nature 8 publications, 29.63%
Elsevier	Elsevier, 6, 22.22% Elsevier 6 publications, 22.22%
IEEE	IEEE, 3, 11.11% IEEE 3 publications, 11.11%
Multidisciplinary Digital Publishing Institute (MDPI)	Multidisciplinary Digital Publishing Institute (MDPI), 2, 7.41% Multidisciplinary Digital Publishing Institute (MDPI) 2 publications, 7.41%
Taylor & Francis	Taylor & Francis, 1, 3.7% Taylor & Francis 1 publication, 3.7%
American Chemical Society (ACS)	American Chemical Society (ACS), 1, 3.7% American Chemical Society (ACS) 1 publication, 3.7%
IOS Press	IOS Press, 1, 3.7% IOS Press 1 publication, 3.7%
	1 2 3 4 5 6 7 8

We do not take into account publications that without a DOI.
Statistics recalculated only for publications connected to researchers, organizations and labs registered on the platform.
Statistics recalculated weekly.

{"yearsCitations":{"type":"bar","data":{"show":true,"labels":[2022,2023,2024],"ids":[0,0,0],"codes":[0,0,0],"imageUrls":["","",""],"datasets":[{"label":"Citations number","data":[9,15,1],"backgroundColor":["#3B82F6","#3B82F6","#3B82F6"],"percentage":["33.33","55.56","3.7"],"barThickness":null}]},"options":{"indexAxis":"x","maintainAspectRatio":true,"scales":{"y":{"ticks":{"precision":0,"autoSkip":false,"font":{"family":"Montserrat"},"color":"#000000"}},"x":{"ticks":{"stepSize":1,"precision":0,"font":{"family":"Montserrat"},"color":"#000000"}}},"plugins":{"legend":{"position":"top","labels":{"font":{"family":"Montserrat"},"color":"#000000"}},"title":{"display":true,"text":"Citations per year","font":{"size":24,"family":"Montserrat","weight":600},"color":"#000000"}}}},"journals":{"type":"bar","data":{"show":true,"labels":["Lecture Notes in Computer Science","Procedia Computer Science","Computers and Geosciences","Water (Switzerland)","Systems Science and Control Engineering","Journal of Personalized Medicine","Knowledge-Based Systems","Results in Engineering","IFIP Advances in Information and Communication Technology","Advances in Intelligent Systems and Computing","ACS applied materials & interfaces","AI Communications","Annals of Operations Research","Lecture Notes in Networks and Systems","Journal of Information and Intelligence"],"ids":[1022,14281,6999,24405,2346,2479,5293,10526,22722,13998,1458,16949,2256,17269,35611],"codes":[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0],"imageUrls":["\/storage\/images\/resized\/voXLqlsvTwv5p3iMQ8Dhs95nqB4AXOG7Taj7G4ra_medium.webp","\/storage\/images\/resized\/GDnYOu1UpMMfMMRV6Aqle4H0YLLsraeD9IP9qScG_medium.webp","\/storage\/images\/resized\/GDnYOu1UpMMfMMRV6Aqle4H0YLLsraeD9IP9qScG_medium.webp","\/storage\/images\/resized\/MjH1ITP7lMYGxeqUZfkt2BnVLgjkk413jwBV97XX_medium.webp","\/storage\/images\/resized\/5YZtvLvkPZuc2JHOaZsjCvGSHFCuC3drUwN3YAc5_medium.webp","\/storage\/images\/resized\/MjH1ITP7lMYGxeqUZfkt2BnVLgjkk413jwBV97XX_medium.webp","\/storage\/images\/resized\/GDnYOu1UpMMfMMRV6Aqle4H0YLLsraeD9IP9qScG_medium.webp","\/storage\/images\/resized\/GDnYOu1UpMMfMMRV6Aqle4H0YLLsraeD9IP9qScG_medium.webp","\/storage\/images\/resized\/voXLqlsvTwv5p3iMQ8Dhs95nqB4AXOG7Taj7G4ra_medium.webp","\/storage\/images\/resized\/voXLqlsvTwv5p3iMQ8Dhs95nqB4AXOG7Taj7G4ra_medium.webp","\/storage\/images\/resized\/iLiQsFqFaSEx6chlGQ5fbAwF6VYU3WWa08hkss0g_medium.webp","\/storage\/images\/resized\/KqWdDIHwzps3KNMz2kSZBC4SgMnkL3bwEKJqtZ2u_medium.webp","\/storage\/images\/resized\/voXLqlsvTwv5p3iMQ8Dhs95nqB4AXOG7Taj7G4ra_medium.webp","\/storage\/images\/resized\/voXLqlsvTwv5p3iMQ8Dhs95nqB4AXOG7Taj7G4ra_medium.webp","\/storage\/images\/resized\/GDnYOu1UpMMfMMRV6Aqle4H0YLLsraeD9IP9qScG_medium.webp"],"datasets":[{"label":"","data":[4,2,1,1,1,1,1,1,1,1,1,1,1,1,1],"backgroundColor":["#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6"],"percentage":[14.81,7.41,3.7,3.7,3.7,3.7,3.7,3.7,3.7,3.7,3.7,3.7,3.7,3.7,3.7],"barThickness":13}]},"options":{"indexAxis":"y","maintainAspectRatio":false,"scales":{"y":{"ticks":{"precision":0,"autoSkip":false,"font":{"family":"Montserrat"},"color":"#000000"}},"x":{"ticks":{"stepSize":null,"precision":0,"font":{"family":"Montserrat"},"color":"#000000"}}},"plugins":{"legend":{"position":"top","labels":{"font":{"family":"Montserrat"},"color":"#000000"}},"title":{"display":true,"text":"Journals","font":{"size":24,"family":"Montserrat","weight":600},"color":"#000000"}}}},"publishers":{"type":"bar","data":{"show":true,"labels":["Springer Nature","Elsevier","IEEE","Multidisciplinary Digital Publishing Institute (MDPI)","Taylor & Francis","American Chemical Society (ACS)","IOS Press"],"ids":[8,17,6953,202,18,40,29],"codes":[0,0,0,0,0,0,0],"imageUrls":["\/storage\/images\/resized\/voXLqlsvTwv5p3iMQ8Dhs95nqB4AXOG7Taj7G4ra_medium.webp","\/storage\/images\/resized\/GDnYOu1UpMMfMMRV6Aqle4H0YLLsraeD9IP9qScG_medium.webp","\/storage\/images\/resized\/6scCJegesojp2jubwY3uKCzTAmgsaH2GIFlg6Hfk_medium.webp","\/storage\/images\/resized\/MjH1ITP7lMYGxeqUZfkt2BnVLgjkk413jwBV97XX_medium.webp","\/storage\/images\/resized\/5YZtvLvkPZuc2JHOaZsjCvGSHFCuC3drUwN3YAc5_medium.webp","\/storage\/images\/resized\/iLiQsFqFaSEx6chlGQ5fbAwF6VYU3WWa08hkss0g_medium.webp","\/storage\/images\/resized\/KqWdDIHwzps3KNMz2kSZBC4SgMnkL3bwEKJqtZ2u_medium.webp"],"datasets":[{"label":"","data":[8,6,3,2,1,1,1],"backgroundColor":["#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6","#3B82F6"],"percentage":[29.63,22.22,11.11,7.41,3.7,3.7,3.7],"barThickness":13}]},"options":{"indexAxis":"y","maintainAspectRatio":false,"scales":{"y":{"ticks":{"precision":0,"autoSkip":false,"font":{"family":"Montserrat"},"color":"#000000"}},"x":{"ticks":{"stepSize":null,"precision":0,"font":{"family":"Montserrat"},"color":"#000000"}}},"plugins":{"legend":{"position":"top","labels":{"font":{"family":"Montserrat"},"color":"#000000"}},"title":{"display":true,"text":"Publishers","font":{"size":24,"family":"Montserrat","weight":600},"color":"#000000"}}}}}

Metrics

Cite this

GOST |

Cite this

GOST Copy

Nikitin N. O. et al. Automated evolutionary approach for the design of composite machine learning pipelines // Future Generation Computer Systems. 2022. Vol. 127. pp. 109-125.

GOST all authors (up to 50) Copy

Nikitin N. O., Vychuzhanin P., Sarafanov M., Polonskaia I. S., Barabanova I. V., Maximov G., Boukhanovsky A., Revin I., Kalyuzhnaya A. V., Kalyuzhnaya A. Automated evolutionary approach for the design of composite machine learning pipelines // Future Generation Computer Systems. 2022. Vol. 127. pp. 109-125.

RIS |

Cite this

RIS Copy

TY - JOUR

DO - 10.1016/j.future.2021.08.022

UR - https://doi.org/10.1016%2Fj.future.2021.08.022

TI - Automated evolutionary approach for the design of composite machine learning pipelines

T2 - Future Generation Computer Systems

AU - Nikitin, Nikolay O

AU - Vychuzhanin, Pavel

AU - Sarafanov, Mikhail

AU - Polonskaia, Iana S

AU - Barabanova, Irina V

AU - Maximov, Gleb

AU - Boukhanovsky, A.

AU - Revin, Ilia

AU - Kalyuzhnaya, Anna V

AU - Kalyuzhnaya, Anna

PY - 2022

DA - 2022/02/01 00:00:00

PB - Elsevier

SP - 109-125

VL - 127

SN - 0167-739X

ER -

BibTex

Cite this

BibTex Copy

@article{2022_Nikitin,

author = {Nikolay O Nikitin and Pavel Vychuzhanin and Mikhail Sarafanov and Iana S Polonskaia and Irina V Barabanova and Gleb Maximov and A. Boukhanovsky and Ilia Revin and Anna V Kalyuzhnaya and Anna Kalyuzhnaya},

title = {Automated evolutionary approach for the design of composite machine learning pipelines},

journal = {Future Generation Computer Systems},

year = {2022},

volume = {127},

publisher = {Elsevier},

month = {feb},

url = {https://doi.org/10.1016%2Fj.future.2021.08.022},

pages = {109--125},

doi = {10.1016/j.future.2021.08.022}

}

Found error?

Publisher

Elsevier

Journal

Future Generation Computer Systems

Quartile SCImago

Quartile WOS

Impact factor

7.5

ISSN

0167739X (Print)

Labs

Computer modeling of natural Systems (NSS Lab)

Profiles