Adaptive Selection of Helper-Objectives with Reinforcement Learning

Arina Buzdalova; Maxim Buzdalov

doi:10.1109/ICMLA.2012.159

Adaptive Selection of Helper-Objectives with Reinforcement Learning

Arina Buzdalova, Maxim Buzdalov

ITMO University

Research output: Chapter in Book/Report/Conference proceeding › Conference Proceeding (Non-Journal item)

7 Citations (Scopus)

Abstract

In this paper a previously proposed method of choosing auxiliary fitness functions is applied to adaptive selection of helper-objectives. Helper-objectives are used in evolutionary computation to enhance the optimization of the primary objective. The method based on choosing between objectives of a single-objective evolutionary algorithm with reinforcement learning is briefly described. It is tested on a model problem. From the results of the experiment, it can be concluded that the method allows to automatically select the most effective helper-objectives and ignore the ineffective ones. It is also shown that the proposed method outperforms multi-objective evolutionary algorithms, that were used with helper-objectives originally.

Original language	English
Title of host publication	ICMLA '12
Subtitle of host publication	Proceedings of the 2012 11th International Conference on Machine Learning and Applications
Publisher	IEEE Press
Pages	66-67
Number of pages	2
Volume	2
ISBN (Print)	978-1-4673-4651-1
DOIs	https://doi.org/10.1109/ICMLA.2012.159
Publication status	Published - 12 Dec 2012
Externally published	Yes
Event	11th IEEE International Conference on Machine Learning and Applications - Cancan, Mexico Duration: 12 Dec 2012 → 15 Dec 2012

Conference

Conference	11th IEEE International Conference on Machine Learning and Applications
Abbreviated title	ICMLA 2012
Country/Territory	Mexico
City	Cancan
Period	12 Dec 2012 → 15 Dec 2012

Access to Document

10.1109/ICMLA.2012.159

Permanent link

Persistent link

Cite this

@inproceedings{d8a8ab4abab3403c8e66cf00e7904871,

title = "Adaptive Selection of Helper-Objectives with Reinforcement Learning",

abstract = "In this paper a previously proposed method of choosing auxiliary fitness functions is applied to adaptive selection of helper-objectives. Helper-objectives are used in evolutionary computation to enhance the optimization of the primary objective. The method based on choosing between objectives of a single-objective evolutionary algorithm with reinforcement learning is briefly described. It is tested on a model problem. From the results of the experiment, it can be concluded that the method allows to automatically select the most effective helper-objectives and ignore the ineffective ones. It is also shown that the proposed method outperforms multi-objective evolutionary algorithms, that were used with helper-objectives originally.",

author = "Arina Buzdalova and Maxim Buzdalov",

year = "2012",

month = dec,

day = "12",

doi = "10.1109/ICMLA.2012.159",

language = "English",

isbn = "978-1-4673-4651-1",

volume = "2",

pages = "66--67",

booktitle = "ICMLA '12",

publisher = "IEEE Press",

address = "United States of America",

note = "11th IEEE International Conference on Machine Learning and Applications, ICMLA 2012 ; Conference date: 12-12-2012 Through 15-12-2012",

}

Buzdalova, A & Buzdalov, M 2012, Adaptive Selection of Helper-Objectives with Reinforcement Learning. in ICMLA '12: Proceedings of the 2012 11th International Conference on Machine Learning and Applications. vol. 2, IEEE Press, pp. 66-67, 11th IEEE International Conference on Machine Learning and Applications, Cancan, Mexico, 12 Dec 2012. https://doi.org/10.1109/ICMLA.2012.159

TY - GEN

T1 - Adaptive Selection of Helper-Objectives with Reinforcement Learning

AU - Buzdalova, Arina

AU - Buzdalov, Maxim

PY - 2012/12/12

Y1 - 2012/12/12

N2 - In this paper a previously proposed method of choosing auxiliary fitness functions is applied to adaptive selection of helper-objectives. Helper-objectives are used in evolutionary computation to enhance the optimization of the primary objective. The method based on choosing between objectives of a single-objective evolutionary algorithm with reinforcement learning is briefly described. It is tested on a model problem. From the results of the experiment, it can be concluded that the method allows to automatically select the most effective helper-objectives and ignore the ineffective ones. It is also shown that the proposed method outperforms multi-objective evolutionary algorithms, that were used with helper-objectives originally.

AB - In this paper a previously proposed method of choosing auxiliary fitness functions is applied to adaptive selection of helper-objectives. Helper-objectives are used in evolutionary computation to enhance the optimization of the primary objective. The method based on choosing between objectives of a single-objective evolutionary algorithm with reinforcement learning is briefly described. It is tested on a model problem. From the results of the experiment, it can be concluded that the method allows to automatically select the most effective helper-objectives and ignore the ineffective ones. It is also shown that the proposed method outperforms multi-objective evolutionary algorithms, that were used with helper-objectives originally.

UR - https://www.scopus.com/record/display.uri?eid=2-s2.0-84873590563&origin=resultslist

U2 - 10.1109/ICMLA.2012.159

DO - 10.1109/ICMLA.2012.159

M3 - Conference Proceeding (Non-Journal item)

SN - 978-1-4673-4651-1

VL - 2

SP - 66

EP - 67

BT - ICMLA '12

PB - IEEE Press

T2 - 11th IEEE International Conference on Machine Learning and Applications

Y2 - 12 December 2012 through 15 December 2012

ER -

Adaptive Selection of Helper-Objectives with Reinforcement Learning

Abstract

Conference

Access to Document

Permanent link

Other files and links

Fingerprint

Cite this