Bayesian Optimization with Multi-objective Acquisition Function for Bilevel Problems

Vedat Dogan; Steven Prestwich

doi:10.1007/978-3-031-26438-2_32

Bayesian Optimization with Multi-objective Acquisition Function for Bilevel Problems

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

Abstract

A bilevel optimization problem consists of an upper-level and a lower-level optimization problem connected to each other hierarchically. Efficient methods exist for special cases, but in general solving these problems is difficult. Bayesian optimization methods are an interesting approach that speed up search using an acquisition function, and this paper proposes a modified Bayesian approach. It treats the upper-level problem as an expensive black-box function, and uses multiple acquisition functions in a multi-objective manner by exploring the Pareto-front. Experiments on popular bilevel benchmark problems show the advantage of the method.

Original language	English
Title of host publication	Artificial Intelligence and Cognitive Science - 30th Irish Conference, AICS 2022, Revised Selected Papers
Editors	Luca Longo, Ruairi O’Reilly
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	409-422
Number of pages	14
ISBN (Print)	9783031264375
DOIs	https://doi.org/10.1007/978-3-031-26438-2_32
Publication status	Published - 2023
Event	30th Irish Conference on Artificial Intelligence and Cognitive Science, AICS 2022 - Munster, Ireland Duration: 8 Dec 2022 → 9 Dec 2022

Publication series

Name	Communications in Computer and Information Science
Volume	1662 CCIS
ISSN (Print)	1865-0929
ISSN (Electronic)	1865-0937

Conference

Conference	30th Irish Conference on Artificial Intelligence and Cognitive Science, AICS 2022
Country/Territory	Ireland
City	Munster
Period	8/12/22 → 9/12/22

Keywords

Bayesian optimization
Bilevel optimization problems
Multi-objective acquisition
Multi-objective optimization

Access to Document

10.1007/978-3-031-26438-2_32

Cite this

Dogan, V., & Prestwich, S. (2023). Bayesian Optimization with Multi-objective Acquisition Function for Bilevel Problems. In L. Longo, & R. O’Reilly (Eds.), Artificial Intelligence and Cognitive Science - 30th Irish Conference, AICS 2022, Revised Selected Papers (pp. 409-422). (Communications in Computer and Information Science; Vol. 1662 CCIS). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-3-031-26438-2_32

Dogan, Vedat ; Prestwich, Steven. / Bayesian Optimization with Multi-objective Acquisition Function for Bilevel Problems. Artificial Intelligence and Cognitive Science - 30th Irish Conference, AICS 2022, Revised Selected Papers. editor / Luca Longo ; Ruairi O’Reilly. Springer Science and Business Media Deutschland GmbH, 2023. pp. 409-422 (Communications in Computer and Information Science).

@inbook{c78eab724898475c8a9187d49f9792c4,

title = "Bayesian Optimization with Multi-objective Acquisition Function for Bilevel Problems",

abstract = "A bilevel optimization problem consists of an upper-level and a lower-level optimization problem connected to each other hierarchically. Efficient methods exist for special cases, but in general solving these problems is difficult. Bayesian optimization methods are an interesting approach that speed up search using an acquisition function, and this paper proposes a modified Bayesian approach. It treats the upper-level problem as an expensive black-box function, and uses multiple acquisition functions in a multi-objective manner by exploring the Pareto-front. Experiments on popular bilevel benchmark problems show the advantage of the method.",

keywords = "Bayesian optimization, Bilevel optimization problems, Multi-objective acquisition, Multi-objective optimization",

author = "Vedat Dogan and Steven Prestwich",

note = "Publisher Copyright: {\textcopyright} 2023, The Author(s).; 30th Irish Conference on Artificial Intelligence and Cognitive Science, AICS 2022 ; Conference date: 08-12-2022 Through 09-12-2022",

year = "2023",

doi = "10.1007/978-3-031-26438-2\_32",

language = "English",

isbn = "9783031264375",

series = "Communications in Computer and Information Science",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "409--422",

editor = "Luca Longo and Ruairi O{\textquoteright}Reilly",

booktitle = "Artificial Intelligence and Cognitive Science - 30th Irish Conference, AICS 2022, Revised Selected Papers",

address = "Germany",

}

Dogan, V & Prestwich, S 2023, Bayesian Optimization with Multi-objective Acquisition Function for Bilevel Problems. in L Longo & R O’Reilly (eds), Artificial Intelligence and Cognitive Science - 30th Irish Conference, AICS 2022, Revised Selected Papers. Communications in Computer and Information Science, vol. 1662 CCIS, Springer Science and Business Media Deutschland GmbH, pp. 409-422, 30th Irish Conference on Artificial Intelligence and Cognitive Science, AICS 2022, Munster, Ireland, 8/12/22. https://doi.org/10.1007/978-3-031-26438-2_32

Bayesian Optimization with Multi-objective Acquisition Function for Bilevel Problems. / Dogan, Vedat ; Prestwich, Steven.
Artificial Intelligence and Cognitive Science - 30th Irish Conference, AICS 2022, Revised Selected Papers. ed. / Luca Longo; Ruairi O’Reilly. Springer Science and Business Media Deutschland GmbH, 2023. p. 409-422 (Communications in Computer and Information Science; Vol. 1662 CCIS).

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

TY - CHAP

T1 - Bayesian Optimization with Multi-objective Acquisition Function for Bilevel Problems

AU - Dogan, Vedat

AU - Prestwich, Steven

PY - 2023

Y1 - 2023

N2 - A bilevel optimization problem consists of an upper-level and a lower-level optimization problem connected to each other hierarchically. Efficient methods exist for special cases, but in general solving these problems is difficult. Bayesian optimization methods are an interesting approach that speed up search using an acquisition function, and this paper proposes a modified Bayesian approach. It treats the upper-level problem as an expensive black-box function, and uses multiple acquisition functions in a multi-objective manner by exploring the Pareto-front. Experiments on popular bilevel benchmark problems show the advantage of the method.

AB - A bilevel optimization problem consists of an upper-level and a lower-level optimization problem connected to each other hierarchically. Efficient methods exist for special cases, but in general solving these problems is difficult. Bayesian optimization methods are an interesting approach that speed up search using an acquisition function, and this paper proposes a modified Bayesian approach. It treats the upper-level problem as an expensive black-box function, and uses multiple acquisition functions in a multi-objective manner by exploring the Pareto-front. Experiments on popular bilevel benchmark problems show the advantage of the method.

KW - Bayesian optimization

KW - Bilevel optimization problems

KW - Multi-objective acquisition

KW - Multi-objective optimization

UR - https://www.scopus.com/pages/publications/85149959160

U2 - 10.1007/978-3-031-26438-2_32

DO - 10.1007/978-3-031-26438-2_32

M3 - Chapter

AN - SCOPUS:85149959160

SN - 9783031264375

T3 - Communications in Computer and Information Science

SP - 409

EP - 422

BT - Artificial Intelligence and Cognitive Science - 30th Irish Conference, AICS 2022, Revised Selected Papers

A2 - Longo, Luca

A2 - O’Reilly, Ruairi

PB - Springer Science and Business Media Deutschland GmbH

T2 - 30th Irish Conference on Artificial Intelligence and Cognitive Science, AICS 2022

Y2 - 8 December 2022 through 9 December 2022

ER -

Dogan V , Prestwich S. Bayesian Optimization with Multi-objective Acquisition Function for Bilevel Problems. In Longo L, O’Reilly R, editors, Artificial Intelligence and Cognitive Science - 30th Irish Conference, AICS 2022, Revised Selected Papers. Springer Science and Business Media Deutschland GmbH. 2023. p. 409-422. (Communications in Computer and Information Science). doi: 10.1007/978-3-031-26438-2_32