Reinforcement Learning initialization by evolutionary formulation: Application for workflow autoscaling in the Cloud

Scientific workflow execution is usually fulfilled through Cloud Computing, but correct autoscaling techniques are needed for proper performance. Reinforcement Learning (RL) has been used for autoscaling, but presents low performance in early stages. Poor initial performance accumulates over episode...

Full description

Saved in:
Bibliographic Details
Published in:Engineering applications of artificial intelligence Vol. 162; p. 112663
Main Authors: Robino, Luciano, Garí, Yisel, Pacini, Elina, Mateos, Cristian, Yannibelli, Virginia, Monge, David A.
Format: Journal Article
Language:English
Published: Elsevier Ltd 24.12.2025
Subjects:
ISSN:0952-1976
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract Scientific workflow execution is usually fulfilled through Cloud Computing, but correct autoscaling techniques are needed for proper performance. Reinforcement Learning (RL) has been used for autoscaling, but presents low performance in early stages. Poor initial performance accumulates over episodes, making the learning process more expensive, which is critical in the context of Cloud autoscaling. Solutions to this problem are sparse and difficult to generalize. Here, we present Reinforcement Learning Initialization by Evolutionary Formulation (ReLIEF), which uses evolutionary algorithm to generate an initial pre-optimized RL policy, that is later refined via RL. Proposed initilization aims to reduce the accumulated losses in monetary cost and execution time (i.e. makespan) during learning. In this article two prominent evolutionary algorithm are used: Non-dominated Sorting Genetic Algorithm III (NSGA-III) and Improved Decomposition-Based Evolutionary Algorithm (I-DBEA). On the other hand, for Reinforcement Learning only Q-Learning in tabular form is used. Four benchmark workflows were used to validate savings produced by the proposal. In 3 out of 4 workflows analyzed, ReLIEF outperformed baseline RL agents. In the remaining workflow, competitive performance was obtained. [Display omitted] •ReLIEF, an evolutionary algorithm to build preoptimized RL policies for autoscaling.•RL-based autoscaler performance is improved during online learning process.•ReLIEF is tested on 4 workflows, comparing RL metrics and EA variants for insights.
AbstractList Scientific workflow execution is usually fulfilled through Cloud Computing, but correct autoscaling techniques are needed for proper performance. Reinforcement Learning (RL) has been used for autoscaling, but presents low performance in early stages. Poor initial performance accumulates over episodes, making the learning process more expensive, which is critical in the context of Cloud autoscaling. Solutions to this problem are sparse and difficult to generalize. Here, we present Reinforcement Learning Initialization by Evolutionary Formulation (ReLIEF), which uses evolutionary algorithm to generate an initial pre-optimized RL policy, that is later refined via RL. Proposed initilization aims to reduce the accumulated losses in monetary cost and execution time (i.e. makespan) during learning. In this article two prominent evolutionary algorithm are used: Non-dominated Sorting Genetic Algorithm III (NSGA-III) and Improved Decomposition-Based Evolutionary Algorithm (I-DBEA). On the other hand, for Reinforcement Learning only Q-Learning in tabular form is used. Four benchmark workflows were used to validate savings produced by the proposal. In 3 out of 4 workflows analyzed, ReLIEF outperformed baseline RL agents. In the remaining workflow, competitive performance was obtained. [Display omitted] •ReLIEF, an evolutionary algorithm to build preoptimized RL policies for autoscaling.•RL-based autoscaler performance is improved during online learning process.•ReLIEF is tested on 4 workflows, comparing RL metrics and EA variants for insights.
ArticleNumber 112663
Author Yannibelli, Virginia
Robino, Luciano
Mateos, Cristian
Monge, David A.
Pacini, Elina
Garí, Yisel
Author_xml – sequence: 1
  givenname: Luciano
  orcidid: 0009-0002-2110-6026
  surname: Robino
  fullname: Robino, Luciano
  email: lrobino@conicet.gov.ar
  organization: ISISTAN-UNICEN-CONICET, Tandil, Buenos Aires, Argentina
– sequence: 2
  givenname: Yisel
  orcidid: 0000-0003-4353-7257
  surname: Garí
  fullname: Garí, Yisel
  email: ygari@uncu.edu.ar
  organization: Laboratorio de Sistemas Inteligentes (LABSIN) - Facultad de Ingeniería, UNCuyo, Mendoza, Argentina
– sequence: 3
  givenname: Elina
  orcidid: 0000-0003-2882-766X
  surname: Pacini
  fullname: Pacini, Elina
  email: epacini@uncu.edu.ar
  organization: Laboratorio de Sistemas Inteligentes (LABSIN) - Facultad de Ingeniería, UNCuyo, Mendoza, Argentina
– sequence: 4
  givenname: Cristian
  orcidid: 0000-0001-5761-1898
  surname: Mateos
  fullname: Mateos, Cristian
  email: cristian.mateos@isistan.unicen.edu.ar
  organization: ISISTAN-UNICEN-CONICET, Tandil, Buenos Aires, Argentina
– sequence: 5
  givenname: Virginia
  orcidid: 0000-0001-7854-7610
  surname: Yannibelli
  fullname: Yannibelli, Virginia
  email: virginia.yannibelli@isistan.unicen.edu.ar
  organization: ISISTAN-UNICEN-CONICET, Tandil, Buenos Aires, Argentina
– sequence: 6
  givenname: David A.
  orcidid: 0000-0001-6444-4610
  surname: Monge
  fullname: Monge, David A.
  email: dmonge@uncu.edu.ar
  organization: Universidad Nacional de Cuyo, Mendoza, Argentina
BookMark eNqFkF1LwzAUhnMxwW36FyR_oDVJ23T1yjH8goEgeh3S9GRmZklJ04356-1WvfbqcHjP-3B4ZmjivAOEbihJKaH8dpuC28i2lSZlhBUppYzzbIKmpCpYQquSX6JZ120JIdki51MU38A47YOCHbiI1yCDM26DjTPRSGu-ZTTe4fqIYe9tf1pkOOKhsevtObvDy7a1Ro2HQ4APPnxp6w9Y9tF3aqCcgTh-Al5Z3zdX6EJL28H175yjj8eH99Vzsn59elkt14linMaklFxr1QBv2EKVFVFZyRZMMSg4lZDTRjMGtOKgmKzrmmdUlXVBqkZXOfCszuaIj1wVfNcF0KINZjf8LygRJ19iK_58iZMvMfoaivdjEYbv9gaC6JQBp6AxAVQUjTf_IX4AP8Z_nQ
Cites_doi 10.1016/j.engappai.2021.104288
10.35833/MPCE.2019.000055
10.1016/j.swevo.2018.03.011
10.23919/INM.2017.7987304
10.1504/IJGUC.2019.099667
10.1109/ACCESS.2022.3149955
10.1371/journal.pone.0279438
10.1016/j.ascom.2018.04.002
10.1016/j.nucengdes.2023.112423
10.1016/j.future.2021.09.007
10.1016/j.swevo.2024.101517
10.1002/pmic.201800489
10.1080/17538947.2015.1130087
10.1016/j.future.2024.04.014
10.1007/s11227-020-03364-1
10.1109/TNSM.2022.3210211
10.1109/TEVC.2014.2339823
10.1145/3603532
10.14778/1920841.1920902
10.1109/TEVC.2013.2281535
10.1016/j.jnca.2019.102464
ContentType Journal Article
Copyright 2025 Elsevier Ltd
Copyright_xml – notice: 2025 Elsevier Ltd
DBID AAYXX
CITATION
DOI 10.1016/j.engappai.2025.112663
DatabaseName CrossRef
DatabaseTitle CrossRef
DatabaseTitleList
DeliveryMethod fulltext_linktorsrc
Discipline Applied Sciences
Computer Science
ExternalDocumentID 10_1016_j_engappai_2025_112663
S0952197625026946
GroupedDBID --K
--M
.DC
.~1
0R~
1B1
1~.
1~5
4.4
457
4G.
5GY
5VS
7-5
71M
8P~
9JN
AABNK
AAEDT
AAEDW
AAIKJ
AAKOC
AALRI
AAOAW
AAQFI
AATTM
AAXKI
AAXUO
AAYFN
AAYWO
ABBOA
ABJNI
ABMAC
ACDAQ
ACGFS
ACLOT
ACRLP
ACVFH
ACZNC
ADBBV
ADCNI
ADEZE
ADTZH
AEBSH
AECPX
AEIPS
AEKER
AENEX
AEUPX
AFJKZ
AFPUW
AFTJW
AGHFR
AGUBO
AGYEJ
AHHHB
AHJVU
AHZHX
AIALX
AIEXJ
AIGII
AIIUN
AIKHN
AITUG
AKBMS
AKRWK
AKYEP
ALMA_UNASSIGNED_HOLDINGS
AMRAJ
ANKPU
AOUOD
APXCP
AXJTR
BJAXD
BKOJK
BLXMC
CS3
DU5
EBS
EFJIC
EFKBS
EFLBG
EO8
EO9
EP2
EP3
F5P
FDB
FIRID
FNPLU
FYGXN
G-Q
GBLVA
GBOLZ
IHE
J1W
JJJVA
KOM
LG9
LY7
M41
MO0
N9A
O-L
O9-
OAUVE
OZT
P-8
P-9
P2P
PC.
Q38
ROL
RPZ
SDF
SDG
SDP
SES
SEW
SPC
SPCBC
SST
SSV
SSZ
T5K
TN5
~G-
~HD
29G
9DU
AAQXK
AAYXX
ABWVN
ABXDB
ACNNM
ACRPL
ADJOM
ADMUD
ADNMO
AGQPQ
ASPBG
AVWKF
AZFZN
CITATION
EJD
FEDTE
FGOYB
G-2
HLZ
HVGLF
HZ~
R2-
SBC
SET
UHS
WUQ
ZMT
ID FETCH-LOGICAL-c261t-7a6ffcde6d28c790c37282c2e561ae41df22e196ec2abbb631c7b509df94e63b3
ISICitedReferencesCount 0
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001597416300012&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0952-1976
IngestDate Sat Nov 29 06:50:35 EST 2025
Sat Nov 29 17:09:13 EST 2025
IsPeerReviewed true
IsScholarly true
Keywords Cloud computing
Workflow
Q-Learning
Improved Decomposition-Based Evolutionary Algorithm
Non-dominated Sorting Genetic Algorithm III
Autoscaling
Language English
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c261t-7a6ffcde6d28c790c37282c2e561ae41df22e196ec2abbb631c7b509df94e63b3
ORCID 0000-0003-2882-766X
0000-0001-5761-1898
0000-0001-7854-7610
0009-0002-2110-6026
0000-0001-6444-4610
0000-0003-4353-7257
ParticipantIDs crossref_primary_10_1016_j_engappai_2025_112663
elsevier_sciencedirect_doi_10_1016_j_engappai_2025_112663
PublicationCentury 2000
PublicationDate 2025-12-24
PublicationDateYYYYMMDD 2025-12-24
PublicationDate_xml – month: 12
  year: 2025
  text: 2025-12-24
  day: 24
PublicationDecade 2020
PublicationTitle Engineering applications of artificial intelligence
PublicationYear 2025
Publisher Elsevier Ltd
Publisher_xml – name: Elsevier Ltd
References Li, Wang, Fan, Geng (b17) 2022; 17
Asafuddoula, Ray, Sarker (b2) 2015; 19
Tournaire, Castel-Taleb, Hyon (b30) 2023; 8
Garí, Monge, Mateos, García Garino (b10) 2019
Drugan (b5) 2019; 44
Yannibelli, Pacini, Monge, Mateos, Rodriguez (b34) 2020
Ali, Keerio, Laghari (b1) 2021; 9
Meade, Fluke (b20) 2018; 23
Song, Wu, Guo, Yan, Suganthan, Zhang, Pedrycz, Das, Mallipeddi, Ajani, Feng (b28) 2024; 86
Monge, Pacini, Mateos, Alba, García Garino (b21) 2020; 149
Asghari, Sohrabi, Yaghmaee (b3) 2021; 77
Yannibelli, Pacini, Monge, Mateos, Rodriguez, Millán, Santos (b35) 2023
Ericson, Mohammadian, Santana (b8) 2017
MacGlashan (b19) 2016
Schad, Dittrich, Quiané-Ruiz (b25) 2010; 3
Shen, Chen, Ma, Zhang (b26) 2025
Saltelli, Ratto, Andres, Campolongo, Cariboni, Gatelli, Saisana, Tarantola (b24) 2008
Hadka (b14) 2024
Silva Filho, M.C., Oliveira, R.L., Monteiro, C.C., Inácio, P.R.M., Freire, M.M., 2017. Cloudsim plus: A cloud computing simulation framework pursuing software engineering principles for improved modularity, extensibility and correctness. In: 2017 IFIP/IEEE Symposium on Integrated Network and Service Management. IM, pp. 400–406.
Liu, Feld, Xue, Garcke, Soddemann, Pan (b18) 2016; 9
Radaideh, Du, Seurin, Seyler, Gu, Wang, Shirvan (b23) 2023; 412
Wei, Kudenko, Liu, Pan, Wu, Meng (b32) 2019
Khadka, Tumer (b15) 2018
Garí, Monge, Mateos (b9) 2022; 127
Xu, Song, Ilager, Gill, Zhao, Ye, Xu (b33) 2022; 19
Netto, Calheiros, Rodrigues, Cunha, Buyya (b22) 2018; 51
Dutreilh, X., Kirgizov, S., 2011. Using reinforcement learning for autonomic resource allocation in clouds: towards a fully automated workflow. In: 7th International Conference on Autonomic and Autonomous Systems. pp. 67–74.
Garí, Pacini, Robino, Mateos, Monge (b12) 2024; 157
Guyon, Orgerie, Morin, Agarwal (b13) 2019; 10
Vandenbrouck, Christiany, Combes, Loux, Brun (b31) 2019; 19
Garí, Monge, Pacini, Mateos, García Garino (b11) 2021; 102
Sutton, Barto (b29) 2018
Kruekaew, Kimpan (b16) 2022; 10
Deb, Jain (b4) 2014; 18
Eiben, Smith (b7) 2015
Guyon (10.1016/j.engappai.2025.112663_b13) 2019; 10
Garí (10.1016/j.engappai.2025.112663_b9) 2022; 127
Khadka (10.1016/j.engappai.2025.112663_b15) 2018
Ali (10.1016/j.engappai.2025.112663_b1) 2021; 9
Xu (10.1016/j.engappai.2025.112663_b33) 2022; 19
Liu (10.1016/j.engappai.2025.112663_b18) 2016; 9
Yannibelli (10.1016/j.engappai.2025.112663_b35) 2023
Garí (10.1016/j.engappai.2025.112663_b10) 2019
Garí (10.1016/j.engappai.2025.112663_b12) 2024; 157
Saltelli (10.1016/j.engappai.2025.112663_b24) 2008
10.1016/j.engappai.2025.112663_b27
Li (10.1016/j.engappai.2025.112663_b17) 2022; 17
Deb (10.1016/j.engappai.2025.112663_b4) 2014; 18
Drugan (10.1016/j.engappai.2025.112663_b5) 2019; 44
MacGlashan (10.1016/j.engappai.2025.112663_b19) 2016
Schad (10.1016/j.engappai.2025.112663_b25) 2010; 3
10.1016/j.engappai.2025.112663_b6
Eiben (10.1016/j.engappai.2025.112663_b7) 2015
Garí (10.1016/j.engappai.2025.112663_b11) 2021; 102
Meade (10.1016/j.engappai.2025.112663_b20) 2018; 23
Sutton (10.1016/j.engappai.2025.112663_b29) 2018
Wei (10.1016/j.engappai.2025.112663_b32) 2019
Ericson (10.1016/j.engappai.2025.112663_b8) 2017
Vandenbrouck (10.1016/j.engappai.2025.112663_b31) 2019; 19
Asafuddoula (10.1016/j.engappai.2025.112663_b2) 2015; 19
Kruekaew (10.1016/j.engappai.2025.112663_b16) 2022; 10
Monge (10.1016/j.engappai.2025.112663_b21) 2020; 149
Shen (10.1016/j.engappai.2025.112663_b26) 2025
Tournaire (10.1016/j.engappai.2025.112663_b30) 2023; 8
Song (10.1016/j.engappai.2025.112663_b28) 2024; 86
Radaideh (10.1016/j.engappai.2025.112663_b23) 2023; 412
Hadka (10.1016/j.engappai.2025.112663_b14) 2024
Asghari (10.1016/j.engappai.2025.112663_b3) 2021; 77
Yannibelli (10.1016/j.engappai.2025.112663_b34) 2020
Netto (10.1016/j.engappai.2025.112663_b22) 2018; 51
References_xml – year: 2024
  ident: b14
  article-title: Moea framework: A free and open source java framework for multiobjective optimization (version 4.5)
– volume: 17
  year: 2022
  ident: b17
  article-title: A novel q-learning algorithm based on improved whale optimization algorithm for path planning
  publication-title: PLoS ONE
– year: 2015
  ident: b7
  publication-title: Introduction To Evolutionary Computing
– volume: 51
  year: 2018
  ident: b22
  article-title: Hpc cloud for scientific and business applications: Taxonomy, vision, and research challenges
  publication-title: ACM Comput. Surv.
– volume: 412
  year: 2023
  ident: b23
  article-title: Neorl: Neuroevolution optimization with reinforcement learning—applications to carbon-free energy systems
  publication-title: Nucl. Eng. Des.
– year: 2008
  ident: b24
  article-title: Global Sensitivity Analysis: The Primer
– volume: 127
  start-page: 168
  year: 2022
  end-page: 180
  ident: b9
  article-title: A q-learning approach for the autoscaling of scientific workflows in the cloud
  publication-title: Future Gener. Comput. Syst.
– volume: 19
  start-page: 3995
  year: 2022
  end-page: 4009
  ident: b33
  article-title: Coscal: Multifaceted scaling of microservices with reinforcement learning
  publication-title: IEEE Trans. Netw. Serv. Manag.
– reference: Dutreilh, X., Kirgizov, S., 2011. Using reinforcement learning for autonomic resource allocation in clouds: towards a fully automated workflow. In: 7th International Conference on Autonomic and Autonomous Systems. pp. 67–74.
– year: 2019
  ident: b10
  article-title: Learning budget assignment policies for autoscaling scientific workflows in the cloud
  publication-title: Clust. Comput.
– volume: 19
  start-page: 445
  year: 2015
  end-page: 460
  ident: b2
  article-title: A decomposition-based evolutionary algorithm for many objective optimization
  publication-title: IEEE Trans. Evol. Comput.
– volume: 23
  start-page: 124
  year: 2018
  end-page: 140
  ident: b20
  article-title: Evaluating virtual hosted desktops for graphics-intensive astronomy
  publication-title: Astron. Comput.
– volume: 86
  year: 2024
  ident: b28
  article-title: Reinforcement learning-assisted evolutionary algorithm: A survey and research opportunities
  publication-title: Swarm Evol. Comput.
– volume: 44
  start-page: 228
  year: 2019
  end-page: 246
  ident: b5
  article-title: Reinforcement learning versus evolutionary computation: A survey on hybrid algorithms
  publication-title: Swarm Evol. Comput.
– volume: 149
  year: 2020
  ident: b21
  article-title: CMI: An online multi-objective genetic autoscaler for scientific and engineering workflows in cloud infrastructures with unreliable virtual machines
  publication-title: J. Netw. Comput. Appl.
– volume: 18
  start-page: 577
  year: 2014
  end-page: 601
  ident: b4
  article-title: An evolutionary many-objective optimization algorithm using reference-point-based non-dominated sorting approach, part i: solving problems with box constraints
  publication-title: IEEE Trans. Evol. Comput.
– year: 2018
  ident: b29
  article-title: Reinforcement Learning: An Introduction
– volume: 157
  start-page: 573
  year: 2024
  end-page: 586
  ident: b12
  article-title: Online rl-based cloud autoscaling for scientific workflows: Evaluation of q-learning and sarsa
  publication-title: Future Gener. Comput. Syst.
– start-page: 249
  year: 2020
  end-page: 263
  ident: b34
  article-title: An nsga-iii-based multi-objective intelligent autoscaler for executing engineering applications in cloud infrastructures
  publication-title: Advances in Soft Computing
– volume: 8
  year: 2023
  ident: b30
  article-title: Efficient computation of optimal thresholds in cloud auto-scaling systems
  publication-title: ACM Trans. Model. Perform. Eval. Comput. Syst.
– volume: 10
  start-page: 17803
  year: 2022
  end-page: 17818
  ident: b16
  article-title: Multi-objective task scheduling optimization for load balancing in cloud computing environment using hybrid artificial bee colony algorithm with reinforcement learning
  publication-title: IEEE Access
– year: 2018
  ident: b15
  article-title: Evolution-guided policy gradient in reinforcement learning
  publication-title: Advances in Neural Information Processing Systems
– volume: 19
  year: 2019
  ident: b31
  article-title: Bioinformatics tools and workflow to select blood biomarkers for early cancer diagnosis: an application to pancreatic cancer
  publication-title: Proteomics
– volume: 102
  year: 2021
  ident: b11
  article-title: Reinforcement learning-based application autoscaling in the cloud: A survey
  publication-title: Eng. Appl. Artif. Intell.
– reference: Silva Filho, M.C., Oliveira, R.L., Monteiro, C.C., Inácio, P.R.M., Freire, M.M., 2017. Cloudsim plus: A cloud computing simulation framework pursuing software engineering principles for improved modularity, extensibility and correctness. In: 2017 IFIP/IEEE Symposium on Integrated Network and Service Management. IM, pp. 400–406.
– start-page: 3
  year: 2025
  end-page: 18
  ident: b26
  article-title: Cost-aware dynamic cloud workflow scheduling using self-attention and evolutionary reinforcement learnin
  publication-title: Service-Oriented Computing
– year: 2016
  ident: b19
  article-title: Brown-umbc reinforcement learning and planning burlap
– start-page: 1
  year: 2019
  end-page: 11
  ident: b32
  article-title: A reinforcement learning based auto-scaling approach for saas providers in dynamic cloud environment
  publication-title: Math. Probl. Eng.
– volume: 10
  start-page: 272
  year: 2019
  end-page: 282
  ident: b13
  article-title: Involving users in energy conservation: A case study in scientific clouds
  publication-title: Int. J. Grid Util. Comput.
– volume: 3
  start-page: 460
  year: 2010
  end-page: 471
  ident: b25
  article-title: Runtime measurements in the cloud: observing, analyzing, and reducing variance
  publication-title: Proc. VLDB Endow.
– volume: 9
  start-page: 748
  year: 2016
  end-page: 765
  ident: b18
  article-title: An efficient geosciences workflow on multi-core processors and gpus: a case study for aerosol optical depth retrieval from modis satellite data
  publication-title: Int. J. Digit. Earth
– volume: 9
  start-page: 404
  year: 2021
  end-page: 415
  ident: b1
  article-title: Optimal site and size of distributed generation allocation in radial distribution network using multi-objective optimization
  publication-title: J. Mod. Power Syst. Clean Energy
– volume: 77
  start-page: 2800
  year: 2021
  end-page: 2828
  ident: b3
  article-title: Task scheduling, resource provisioning, and load balancing on scientific workflows using parallel SARSA reinforcement learning agents and genetic algorithm
  publication-title: J. Supercomput.
– start-page: 308
  year: 2017
  end-page: 314
  ident: b8
  article-title: Analysis of performance variability in public cloud computing
  publication-title: Proceedings of the 2017 IEEE International Conference on Information Reuse and Integration
– year: 2023
  ident: b35
  article-title: An in-Depth Benchmarking of Evolutionary and Swarm Intelligence Algorithms for Autoscaling Parameter Sweep Applications on Public Clouds
– volume: 102
  year: 2021
  ident: 10.1016/j.engappai.2025.112663_b11
  article-title: Reinforcement learning-based application autoscaling in the cloud: A survey
  publication-title: Eng. Appl. Artif. Intell.
  doi: 10.1016/j.engappai.2021.104288
– volume: 9
  start-page: 404
  year: 2021
  ident: 10.1016/j.engappai.2025.112663_b1
  article-title: Optimal site and size of distributed generation allocation in radial distribution network using multi-objective optimization
  publication-title: J. Mod. Power Syst. Clean Energy
  doi: 10.35833/MPCE.2019.000055
– volume: 44
  start-page: 228
  year: 2019
  ident: 10.1016/j.engappai.2025.112663_b5
  article-title: Reinforcement learning versus evolutionary computation: A survey on hybrid algorithms
  publication-title: Swarm Evol. Comput.
  doi: 10.1016/j.swevo.2018.03.011
– ident: 10.1016/j.engappai.2025.112663_b27
  doi: 10.23919/INM.2017.7987304
– volume: 10
  start-page: 272
  year: 2019
  ident: 10.1016/j.engappai.2025.112663_b13
  article-title: Involving users in energy conservation: A case study in scientific clouds
  publication-title: Int. J. Grid Util. Comput.
  doi: 10.1504/IJGUC.2019.099667
– volume: 10
  start-page: 17803
  year: 2022
  ident: 10.1016/j.engappai.2025.112663_b16
  article-title: Multi-objective task scheduling optimization for load balancing in cloud computing environment using hybrid artificial bee colony algorithm with reinforcement learning
  publication-title: IEEE Access
  doi: 10.1109/ACCESS.2022.3149955
– volume: 17
  year: 2022
  ident: 10.1016/j.engappai.2025.112663_b17
  article-title: A novel q-learning algorithm based on improved whale optimization algorithm for path planning
  publication-title: PLoS ONE
  doi: 10.1371/journal.pone.0279438
– volume: 23
  start-page: 124
  year: 2018
  ident: 10.1016/j.engappai.2025.112663_b20
  article-title: Evaluating virtual hosted desktops for graphics-intensive astronomy
  publication-title: Astron. Comput.
  doi: 10.1016/j.ascom.2018.04.002
– volume: 412
  year: 2023
  ident: 10.1016/j.engappai.2025.112663_b23
  article-title: Neorl: Neuroevolution optimization with reinforcement learning—applications to carbon-free energy systems
  publication-title: Nucl. Eng. Des.
  doi: 10.1016/j.nucengdes.2023.112423
– year: 2008
  ident: 10.1016/j.engappai.2025.112663_b24
– volume: 127
  start-page: 168
  year: 2022
  ident: 10.1016/j.engappai.2025.112663_b9
  article-title: A q-learning approach for the autoscaling of scientific workflows in the cloud
  publication-title: Future Gener. Comput. Syst.
  doi: 10.1016/j.future.2021.09.007
– year: 2018
  ident: 10.1016/j.engappai.2025.112663_b15
  article-title: Evolution-guided policy gradient in reinforcement learning
– year: 2018
  ident: 10.1016/j.engappai.2025.112663_b29
– start-page: 249
  year: 2020
  ident: 10.1016/j.engappai.2025.112663_b34
  article-title: An nsga-iii-based multi-objective intelligent autoscaler for executing engineering applications in cloud infrastructures
– year: 2015
  ident: 10.1016/j.engappai.2025.112663_b7
– year: 2023
  ident: 10.1016/j.engappai.2025.112663_b35
– volume: 86
  year: 2024
  ident: 10.1016/j.engappai.2025.112663_b28
  article-title: Reinforcement learning-assisted evolutionary algorithm: A survey and research opportunities
  publication-title: Swarm Evol. Comput.
  doi: 10.1016/j.swevo.2024.101517
– volume: 19
  year: 2019
  ident: 10.1016/j.engappai.2025.112663_b31
  article-title: Bioinformatics tools and workflow to select blood biomarkers for early cancer diagnosis: an application to pancreatic cancer
  publication-title: Proteomics
  doi: 10.1002/pmic.201800489
– year: 2019
  ident: 10.1016/j.engappai.2025.112663_b10
  article-title: Learning budget assignment policies for autoscaling scientific workflows in the cloud
  publication-title: Clust. Comput.
– volume: 51
  year: 2018
  ident: 10.1016/j.engappai.2025.112663_b22
  article-title: Hpc cloud for scientific and business applications: Taxonomy, vision, and research challenges
  publication-title: ACM Comput. Surv.
– start-page: 3
  year: 2025
  ident: 10.1016/j.engappai.2025.112663_b26
  article-title: Cost-aware dynamic cloud workflow scheduling using self-attention and evolutionary reinforcement learnin
– start-page: 1
  year: 2019
  ident: 10.1016/j.engappai.2025.112663_b32
  article-title: A reinforcement learning based auto-scaling approach for saas providers in dynamic cloud environment
  publication-title: Math. Probl. Eng.
– volume: 9
  start-page: 748
  year: 2016
  ident: 10.1016/j.engappai.2025.112663_b18
  article-title: An efficient geosciences workflow on multi-core processors and gpus: a case study for aerosol optical depth retrieval from modis satellite data
  publication-title: Int. J. Digit. Earth
  doi: 10.1080/17538947.2015.1130087
– year: 2024
  ident: 10.1016/j.engappai.2025.112663_b14
– volume: 157
  start-page: 573
  year: 2024
  ident: 10.1016/j.engappai.2025.112663_b12
  article-title: Online rl-based cloud autoscaling for scientific workflows: Evaluation of q-learning and sarsa
  publication-title: Future Gener. Comput. Syst.
  doi: 10.1016/j.future.2024.04.014
– volume: 77
  start-page: 2800
  year: 2021
  ident: 10.1016/j.engappai.2025.112663_b3
  article-title: Task scheduling, resource provisioning, and load balancing on scientific workflows using parallel SARSA reinforcement learning agents and genetic algorithm
  publication-title: J. Supercomput.
  doi: 10.1007/s11227-020-03364-1
– volume: 19
  start-page: 3995
  year: 2022
  ident: 10.1016/j.engappai.2025.112663_b33
  article-title: Coscal: Multifaceted scaling of microservices with reinforcement learning
  publication-title: IEEE Trans. Netw. Serv. Manag.
  doi: 10.1109/TNSM.2022.3210211
– volume: 19
  start-page: 445
  year: 2015
  ident: 10.1016/j.engappai.2025.112663_b2
  article-title: A decomposition-based evolutionary algorithm for many objective optimization
  publication-title: IEEE Trans. Evol. Comput.
  doi: 10.1109/TEVC.2014.2339823
– volume: 8
  year: 2023
  ident: 10.1016/j.engappai.2025.112663_b30
  article-title: Efficient computation of optimal thresholds in cloud auto-scaling systems
  publication-title: ACM Trans. Model. Perform. Eval. Comput. Syst.
  doi: 10.1145/3603532
– volume: 3
  start-page: 460
  year: 2010
  ident: 10.1016/j.engappai.2025.112663_b25
  article-title: Runtime measurements in the cloud: observing, analyzing, and reducing variance
  publication-title: Proc. VLDB Endow.
  doi: 10.14778/1920841.1920902
– volume: 18
  start-page: 577
  year: 2014
  ident: 10.1016/j.engappai.2025.112663_b4
  article-title: An evolutionary many-objective optimization algorithm using reference-point-based non-dominated sorting approach, part i: solving problems with box constraints
  publication-title: IEEE Trans. Evol. Comput.
  doi: 10.1109/TEVC.2013.2281535
– start-page: 308
  year: 2017
  ident: 10.1016/j.engappai.2025.112663_b8
  article-title: Analysis of performance variability in public cloud computing
– ident: 10.1016/j.engappai.2025.112663_b6
– year: 2016
  ident: 10.1016/j.engappai.2025.112663_b19
– volume: 149
  year: 2020
  ident: 10.1016/j.engappai.2025.112663_b21
  article-title: CMI: An online multi-objective genetic autoscaler for scientific and engineering workflows in cloud infrastructures with unreliable virtual machines
  publication-title: J. Netw. Comput. Appl.
  doi: 10.1016/j.jnca.2019.102464
SSID ssj0003846
Score 2.444218
Snippet Scientific workflow execution is usually fulfilled through Cloud Computing, but correct autoscaling techniques are needed for proper performance. Reinforcement...
SourceID crossref
elsevier
SourceType Index Database
Publisher
StartPage 112663
SubjectTerms Autoscaling
Cloud computing
Improved Decomposition-Based Evolutionary Algorithm
Non-dominated Sorting Genetic Algorithm III
Q-Learning
Workflow
Title Reinforcement Learning initialization by evolutionary formulation: Application for workflow autoscaling in the Cloud
URI https://dx.doi.org/10.1016/j.engappai.2025.112663
Volume 162
WOSCitedRecordID wos001597416300012&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVESC
  databaseName: Elsevier SD Freedom Collection Journals 2021
  issn: 0952-1976
  databaseCode: AIEXJ
  dateStart: 19950201
  customDbUrl:
  isFulltext: true
  dateEnd: 99991231
  titleUrlDefault: https://www.sciencedirect.com
  omitProxy: false
  ssIdentifier: ssj0003846
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1LT9tAEF4F6IELhRYEbUF74IacJuvn9oYCCBCgqqJSbtZ6vUZBkY0SJ9Afx39j9mVvWlTgwMWyNvLG9nzeeew3MwjtFwK-IB70vFAw3wtoL_SSKIk8QkCdBTSH80I1m4ivrpLhkP7sdB5tLsx8HJdl8vBA795V1DAGwpaps28QdzMpDMA5CB2OIHY4vkrwv4QqhspV3M_WT5V5KyP4mMcm7VJanWJu7kMS56Ttajp5qVhhu6-tiIiSvlWMq_sDNqurKcjVpMJIs3UwrmYL7T6dGocH7ga54hxMFDlJtQpxqoG2-z7gqavo7cVMxluqhh_E9J7-kdIZskdUu_nFR6ovlaSolY2WuQQjWlMIB2YdcwMcJJRkEZ1X3UQqidenuk1Ms2ibNVwvuzIPSq-T_2gEHZy47YryBp6YjbryL7rtBYsluP9SjQ1h0XLhblM7TyrnSfU8S2iFxCEFvbByeHY8PG9MAT_RmWL2CZwU9efv6HnryLF4rtfRmnFV8KGG2AbqiPIT-mjcFmyUwhSGbGcQO_YZ1QsgxBaEeBGEOPuDXRBiB4Q_sANB-QO2EMQOBGFCDBDECoKb6PfJ8fXg1DP9PTwOfnvtxSwqCp6LKCcJj2mP-zFJCCcCbHomgn5eECJAQwhOWJZlkd_ncQYGbl7QQER-5m-h5bIqxTbCmSAhpwycZREGlLMkDHnEwJXnjOVR4e-g7_a9pne6jEv6f5nuIGpff2qMUW1kpoCsF6798uZ_-4pWW-h_Q8v1ZCZ20Qc-r0fTyZ6B1RPgmrcC
linkProvider Elsevier
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Reinforcement+Learning+initialization+by+evolutionary+formulation%3A+Application+for+workflow+autoscaling+in+the+Cloud&rft.jtitle=Engineering+applications+of+artificial+intelligence&rft.au=Robino%2C+Luciano&rft.au=Gar%C3%AD%2C+Yisel&rft.au=Pacini%2C+Elina&rft.au=Mateos%2C+Cristian&rft.date=2025-12-24&rft.issn=0952-1976&rft.volume=162&rft.spage=112663&rft_id=info:doi/10.1016%2Fj.engappai.2025.112663&rft.externalDBID=n%2Fa&rft.externalDocID=10_1016_j_engappai_2025_112663
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0952-1976&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0952-1976&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0952-1976&client=summon