Toward a new family of hybrid evolutionary algorithms

Lourdes Uribe, Oliver Schütze, Adriana Lara

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearch

Abstract

© Springer Nature Switzerland AG 2019. Multi-objective optimization problems (MOPs) arise in a natural way in diverse knowledge areas. Multi-objective evolutionary algorithms (MOEAs) have been applied successfully to solve this type of optimization problems over the last two decades. However, until now MOEAs need quite a few resources in order to obtain acceptable Pareto set/front approximations. Even more, in certain cases when the search space is highly constrained, MOEAs may have troubles when approximating the solution set. When dealing with constrained MOPs (CMOPs), MOEAs usually apply penalization methods. One possibility to overcome these situations is the hybridization of MOEAs with local search operators. If the local search operator is based on classical mathematical programming, gradient information is used, leading to a relatively high computational cost. In this work, we give an overview of our recently proposed constraint handling methods and their corresponding hybrid algorithms. These methods have specific mechanisms that deal with the constraints in a wiser way without increasing their cost. Both methods do not explicitly compute the gradients but extract this information in the best manner out of the current population of the MOEAs. We conjecture that these techniques will allow for the fast and reliable treatment of CMOPs in the near future. Numerical results indicate that these ideas already yield competitive results in many cases.
Original languageAmerican English
Title of host publicationToward a new family of hybrid evolutionary algorithms
Pages78-90
Number of pages68
ISBN (Electronic)9783030125974
DOIs
StatePublished - 1 Jan 2019
EventLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) -
Duration: 1 Jan 2019 → …

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11411 LNCS
ISSN (Print)0302-9743

Conference

ConferenceLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Period1/01/19 → …

Fingerprint

Hybrid Evolutionary Algorithm
Multi-objective Evolutionary Algorithm
Evolutionary algorithms
Multiobjective Optimization Problems
Multiobjective optimization
Local Search
Mathematical operators
Penalization Method
Gradient
Pareto Set
Constraint Handling
Mathematical programming
Constrained optimization
Constrained Optimization Problem
Hybrid Algorithm
Operator
Set theory
Solution Set
Mathematical Programming
Search Space

Cite this

Uribe, L., Schütze, O., & Lara, A. (2019). Toward a new family of hybrid evolutionary algorithms. In Toward a new family of hybrid evolutionary algorithms (pp. 78-90). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11411 LNCS). https://doi.org/10.1007/978-3-030-12598-1_7
Uribe, Lourdes ; Schütze, Oliver ; Lara, Adriana. / Toward a new family of hybrid evolutionary algorithms. Toward a new family of hybrid evolutionary algorithms. 2019. pp. 78-90 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{25e573a9fd6345bdaedeae90970eb015,
title = "Toward a new family of hybrid evolutionary algorithms",
abstract = "{\circledC} Springer Nature Switzerland AG 2019. Multi-objective optimization problems (MOPs) arise in a natural way in diverse knowledge areas. Multi-objective evolutionary algorithms (MOEAs) have been applied successfully to solve this type of optimization problems over the last two decades. However, until now MOEAs need quite a few resources in order to obtain acceptable Pareto set/front approximations. Even more, in certain cases when the search space is highly constrained, MOEAs may have troubles when approximating the solution set. When dealing with constrained MOPs (CMOPs), MOEAs usually apply penalization methods. One possibility to overcome these situations is the hybridization of MOEAs with local search operators. If the local search operator is based on classical mathematical programming, gradient information is used, leading to a relatively high computational cost. In this work, we give an overview of our recently proposed constraint handling methods and their corresponding hybrid algorithms. These methods have specific mechanisms that deal with the constraints in a wiser way without increasing their cost. Both methods do not explicitly compute the gradients but extract this information in the best manner out of the current population of the MOEAs. We conjecture that these techniques will allow for the fast and reliable treatment of CMOPs in the near future. Numerical results indicate that these ideas already yield competitive results in many cases.",
author = "Lourdes Uribe and Oliver Sch{\"u}tze and Adriana Lara",
year = "2019",
month = "1",
day = "1",
doi = "10.1007/978-3-030-12598-1_7",
language = "American English",
isbn = "9783030125974",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "78--90",
booktitle = "Toward a new family of hybrid evolutionary algorithms",

}

Uribe, L, Schütze, O & Lara, A 2019, Toward a new family of hybrid evolutionary algorithms. in Toward a new family of hybrid evolutionary algorithms. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11411 LNCS, pp. 78-90, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 1/01/19. https://doi.org/10.1007/978-3-030-12598-1_7

Toward a new family of hybrid evolutionary algorithms. / Uribe, Lourdes; Schütze, Oliver; Lara, Adriana.

Toward a new family of hybrid evolutionary algorithms. 2019. p. 78-90 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11411 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearch

TY - GEN

T1 - Toward a new family of hybrid evolutionary algorithms

AU - Uribe, Lourdes

AU - Schütze, Oliver

AU - Lara, Adriana

PY - 2019/1/1

Y1 - 2019/1/1

N2 - © Springer Nature Switzerland AG 2019. Multi-objective optimization problems (MOPs) arise in a natural way in diverse knowledge areas. Multi-objective evolutionary algorithms (MOEAs) have been applied successfully to solve this type of optimization problems over the last two decades. However, until now MOEAs need quite a few resources in order to obtain acceptable Pareto set/front approximations. Even more, in certain cases when the search space is highly constrained, MOEAs may have troubles when approximating the solution set. When dealing with constrained MOPs (CMOPs), MOEAs usually apply penalization methods. One possibility to overcome these situations is the hybridization of MOEAs with local search operators. If the local search operator is based on classical mathematical programming, gradient information is used, leading to a relatively high computational cost. In this work, we give an overview of our recently proposed constraint handling methods and their corresponding hybrid algorithms. These methods have specific mechanisms that deal with the constraints in a wiser way without increasing their cost. Both methods do not explicitly compute the gradients but extract this information in the best manner out of the current population of the MOEAs. We conjecture that these techniques will allow for the fast and reliable treatment of CMOPs in the near future. Numerical results indicate that these ideas already yield competitive results in many cases.

AB - © Springer Nature Switzerland AG 2019. Multi-objective optimization problems (MOPs) arise in a natural way in diverse knowledge areas. Multi-objective evolutionary algorithms (MOEAs) have been applied successfully to solve this type of optimization problems over the last two decades. However, until now MOEAs need quite a few resources in order to obtain acceptable Pareto set/front approximations. Even more, in certain cases when the search space is highly constrained, MOEAs may have troubles when approximating the solution set. When dealing with constrained MOPs (CMOPs), MOEAs usually apply penalization methods. One possibility to overcome these situations is the hybridization of MOEAs with local search operators. If the local search operator is based on classical mathematical programming, gradient information is used, leading to a relatively high computational cost. In this work, we give an overview of our recently proposed constraint handling methods and their corresponding hybrid algorithms. These methods have specific mechanisms that deal with the constraints in a wiser way without increasing their cost. Both methods do not explicitly compute the gradients but extract this information in the best manner out of the current population of the MOEAs. We conjecture that these techniques will allow for the fast and reliable treatment of CMOPs in the near future. Numerical results indicate that these ideas already yield competitive results in many cases.

UR - https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85063056467&origin=inward

UR - https://www.scopus.com/inward/citedby.uri?partnerID=HzOxMe3b&scp=85063056467&origin=inward

U2 - 10.1007/978-3-030-12598-1_7

DO - 10.1007/978-3-030-12598-1_7

M3 - Conference contribution

SN - 9783030125974

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 78

EP - 90

BT - Toward a new family of hybrid evolutionary algorithms

ER -

Uribe L, Schütze O, Lara A. Toward a new family of hybrid evolutionary algorithms. In Toward a new family of hybrid evolutionary algorithms. 2019. p. 78-90. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-030-12598-1_7