A Brief Guide to Multi-Objective Reinforcement Learning and Planning: JAAMAS Track

Conor F. Hayes; Roxana Radulescu; Eugenio Bargiacchi; Johan Källström; Matthew Macfarlane; Mathieu Reymond; Timothy Verstraeten; Luisa Zintgraf; Richard Dazeley; Fredrik Heintz; Enda Howley; Athirai A. Irissappane; Patrick Mannion; Ann Nowe; Gabriel De Oliveira Ramos; Marcello Restelli; Peter Vamplew; Diederik M. Roijers

A Brief Guide to Multi-Objective Reinforcement Learning and Planning: JAAMAS Track

Conor F. Hayes, Roxana Radulescu, Eugenio Bargiacchi, Johan Källström, Matthew Macfarlane, Mathieu Reymond, Timothy Verstraeten, Luisa Zintgraf, Richard Dazeley, Fredrik Heintz, Enda Howley, Athirai A. Irissappane, Patrick Mannion, Ann Nowe, Gabriel De Oliveira Ramos, Marcello Restelli, Peter Vamplew, Diederik M. Roijers

Research output: Chapter in Book/Report/Conference proceeding › Conference paper

Abstract

Real-world sequential decision-making tasks are usually complex, and require trade-offs between multiple -- often conflicting -- objectives. However, the majority of research in reinforcement learning (RL) and decision-theoretic planning assumes a single objective, or that multiple objectives can be handled via a predefined weighted sum over the objectives. Such approaches may oversimplify the underlying problem, and produce suboptimal results.
This extended abstract outlines the limitations of using a semi-blind iterative process to solve multi-objective decision making problems. Our extended paper serves as a guide for the application of explicitly multi-objective methods to difficult problems.

Original language	English
Title of host publication	The 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2023)
Publisher	International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS)
Publication status	Accepted/In press - 2023
Event	The 22nd International Conference on Autonomous Agents and Multiagent Systems - London, United Kingdom Duration: 29 May 2023 → 2 Jun 2023 https://aamas2023.soton.ac.uk

Conference

Conference	The 22nd International Conference on Autonomous Agents and Multiagent Systems
Abbreviated title	AAMAS 2023
Country	United Kingdom
City	London
Period	29/05/23 → 2/06/23
Internet address	https://aamas2023.soton.ac.uk

FWOTM1108: Decision-making in team-reward multi-objective multi-agent domains
Radulescu, R. & Nowe, A.
1/10/22 → 30/09/25
Project: Fundamental
VLAAI1: Subsidie: Onderzoeksprogramma Artificiële Intelligentie (AI) Vlaanderen
Nowe, A. & Vanderborght, B.
1/07/19 → 31/12/23
Project: Applied

Cite this

Hayes, C. F., Radulescu, R., Bargiacchi, E., Källström, J., Macfarlane, M., Reymond, M., Verstraeten, T., Zintgraf, L., Dazeley, R., Heintz, F., Howley, E., Irissappane, A. A., Mannion, P., Nowe, A., De Oliveira Ramos, G., Restelli, M., Vamplew, P., & Roijers, D. M. (Accepted/In press). A Brief Guide to Multi-Objective Reinforcement Learning and Planning: JAAMAS Track. In The 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2023) International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS).

Hayes, Conor F. ; Radulescu, Roxana ; Bargiacchi, Eugenio ; Källström, Johan ; Macfarlane, Matthew ; Reymond, Mathieu ; Verstraeten, Timothy ; Zintgraf, Luisa ; Dazeley, Richard ; Heintz, Fredrik ; Howley, Enda ; Irissappane, Athirai A. ; Mannion, Patrick ; Nowe, Ann ; De Oliveira Ramos, Gabriel ; Restelli, Marcello ; Vamplew, Peter ; Roijers, Diederik M. / A Brief Guide to Multi-Objective Reinforcement Learning and Planning : JAAMAS Track. The 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2023). International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS), 2023.

@inproceedings{6f906192628547bd894428f171bf01ce,

title = "A Brief Guide to Multi-Objective Reinforcement Learning and Planning: JAAMAS Track",

abstract = "Real-world sequential decision-making tasks are usually complex, and require trade-offs between multiple -- often conflicting -- objectives. However, the majority of research in reinforcement learning (RL) and decision-theoretic planning assumes a single objective, or that multiple objectives can be handled via a predefined weighted sum over the objectives. Such approaches may oversimplify the underlying problem, and produce suboptimal results. This extended abstract outlines the limitations of using a semi-blind iterative process to solve multi-objective decision making problems. Our extended paper serves as a guide for the application of explicitly multi-objective methods to difficult problems.",

author = "Hayes, {Conor F.} and Roxana Radulescu and Eugenio Bargiacchi and Johan K{\"a}llstr{\"o}m and Matthew Macfarlane and Mathieu Reymond and Timothy Verstraeten and Luisa Zintgraf and Richard Dazeley and Fredrik Heintz and Enda Howley and Irissappane, {Athirai A.} and Patrick Mannion and Ann Nowe and {De Oliveira Ramos}, Gabriel and Marcello Restelli and Peter Vamplew and Roijers, {Diederik M.}",

year = "2023",

language = "English",

booktitle = "The 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2023)",

publisher = "International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS)",

note = "The 22nd International Conference on Autonomous Agents and Multiagent Systems, AAMAS 2023 ; Conference date: 29-05-2023 Through 02-06-2023",

url = "https://aamas2023.soton.ac.uk",

}

Hayes, CF, Radulescu, R , Bargiacchi, E, Källström, J, Macfarlane, M, Reymond, M , Verstraeten, T, Zintgraf, L, Dazeley, R, Heintz, F, Howley, E, Irissappane, AA, Mannion, P, Nowe, A, De Oliveira Ramos, G, Restelli, M, Vamplew, P & Roijers, DM 2023, A Brief Guide to Multi-Objective Reinforcement Learning and Planning: JAAMAS Track. in The 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2023). International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS), The 22nd International Conference on Autonomous Agents and Multiagent Systems, London, United Kingdom, 29/05/23.

A Brief Guide to Multi-Objective Reinforcement Learning and Planning : JAAMAS Track. / Hayes, Conor F.; Radulescu, Roxana ; Bargiacchi, Eugenio; Källström, Johan; Macfarlane, Matthew; Reymond, Mathieu ; Verstraeten, Timothy; Zintgraf, Luisa; Dazeley, Richard; Heintz, Fredrik; Howley, Enda; Irissappane, Athirai A.; Mannion, Patrick; Nowe, Ann; De Oliveira Ramos, Gabriel; Restelli, Marcello; Vamplew, Peter; Roijers, Diederik M.

The 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2023). International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS), 2023.

Research output: Chapter in Book/Report/Conference proceeding › Conference paper

TY - GEN

T1 - A Brief Guide to Multi-Objective Reinforcement Learning and Planning

T2 - The 22nd International Conference on Autonomous Agents and Multiagent Systems

AU - Hayes, Conor F.

AU - Radulescu, Roxana

AU - Bargiacchi, Eugenio

AU - Källström, Johan

AU - Macfarlane, Matthew

AU - Reymond, Mathieu

AU - Verstraeten, Timothy

AU - Zintgraf, Luisa

AU - Dazeley, Richard

AU - Heintz, Fredrik

AU - Howley, Enda

AU - Irissappane, Athirai A.

AU - Mannion, Patrick

AU - Nowe, Ann

AU - De Oliveira Ramos, Gabriel

AU - Restelli, Marcello

AU - Vamplew, Peter

AU - Roijers, Diederik M.

PY - 2023

Y1 - 2023

N2 - Real-world sequential decision-making tasks are usually complex, and require trade-offs between multiple -- often conflicting -- objectives. However, the majority of research in reinforcement learning (RL) and decision-theoretic planning assumes a single objective, or that multiple objectives can be handled via a predefined weighted sum over the objectives. Such approaches may oversimplify the underlying problem, and produce suboptimal results. This extended abstract outlines the limitations of using a semi-blind iterative process to solve multi-objective decision making problems. Our extended paper serves as a guide for the application of explicitly multi-objective methods to difficult problems.

AB - Real-world sequential decision-making tasks are usually complex, and require trade-offs between multiple -- often conflicting -- objectives. However, the majority of research in reinforcement learning (RL) and decision-theoretic planning assumes a single objective, or that multiple objectives can be handled via a predefined weighted sum over the objectives. Such approaches may oversimplify the underlying problem, and produce suboptimal results. This extended abstract outlines the limitations of using a semi-blind iterative process to solve multi-objective decision making problems. Our extended paper serves as a guide for the application of explicitly multi-objective methods to difficult problems.

M3 - Conference paper

BT - The 22nd International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2023)

PB - International Foundation for Autonomous Agents and Multiagent Systems (IFAAMAS)

Y2 - 29 May 2023 through 2 June 2023

ER -

A Brief Guide to Multi-Objective Reinforcement Learning and Planning: JAAMAS Track

Abstract

Conference

Projects

FWOTM1108: Decision-making in team-reward multi-objective multi-agent domains

VLAAI1: Subsidie: Onderzoeksprogramma Artificiële Intelligentie (AI) Vlaanderen

Cite this