IJCAI-22 tutorial: When Multiple Agents Care About More than One Objective

by Roxana Rădulescu and Diederik M. Roijers

Place and time: July 24 (Sun), Room Lehar 3 at the Wiener Messe, Time: 14h - 15h; 15h30 - 16h

Many real-world decision problems have more than a single objective, and more than one agent. Mathematically, this means that agents receive a reward vector, rather than a scalar reward. This might seem like a minor modelling change. However, it changes the problem to its core; from its optimisation criterion, to its solutions. For example, the well-known game-theory result that every (single-objective) normal form game has a Nash equilibrium, no longer holds when the agents care about more than one objective.

In this tutorial, we will start from what it means to care about more than one aspect of the solution, and why you should care about that when modelling multi-agent settings. Then we will go into what agents should optimise for in multi-objective settings, and discuss different assumptions, culminating in a taxonomy of multi-objective multi-agent settings and accompanying solution concepts. We will then follow up with a few recent and surprising results from the multi-objective decision making field. Finally, we will end with tips and tricks for identifying and dealing with multi-objective multi-agent problems. (They are everywhere!)

Slides and other materials

The slides for this tutorial are available here.

Materials from older multi-objective decision making lectures and tutorials

The slides for the tutorial "Multi-Objective Planning and Reinforcement Learning" by Diederik M. Roijers & Shimon Whiteson (UOxford) on July 10, 2018, at AAMAS conference (FAIM2018) at the Stockholmsmässan in Stockholm can be found [here].

The slides for the guest lecture on "Multi-Objective Sequential Decision Making" (by Diederik M. Roijers) for the Algorithms for Planning and Scheduling course at Delft University of Technology (September 2016) are available here: [slides].

The material for the EASSS 2016 tutorial: "Multi-Objective Decision Making" (MODeM) by Diederik M. Roijers & Shimon Whiteson is available here: [slides]

The slides for the ICAPS 2016 tutorial "Multi-Objective Planning under Uncertainty" by Diederik M. Roijers & Shimon Whiteson are available here: [slides] [YouTube].

This is the material for the IJCAI 2015 tutorial: "T15: Multi-Objective Decision Making" Diederik M. Roijers & Shimon Whiteson: [slides].

The slides for the guest lecture on "Multi-objective decision problems" (by Diederik M. Roijers) for the Advanced Topics in Autonomous Agents course at the University of Amsterdam (April 2015) are available here : [slides].