Jörg Rothe

Bio

Jörg Rothe is professor for computer science at the Heinrich-Heine-Universität Düsseldorf, Germany. His research at the crossroads of economics, computational complexity, and artificial intelligence focuses on collective decision making, computational social choice, algorithmic game theory, fair division of indivisible goods, as well as argumentation theory. He is an associative editor of the Annals of Mathematics and Artificial Intelligence and will be the local organizing chair of Theoretical Aspects of Rationality and Knowledge TARK 2025. Previously, he has been associative editor of the Journal of Artificial Intelligence Research, co-chair of the Third International Workshop on Computational Social Choice COMSOC 2010, program chair of the Fifth International Conference on Algorithmic Decision Theory ADT 2017, and co-chair of the 19th European Conference on Multi-Agent Systems EUMAS 2022.

Title

Economics and Computation: The Second Edition

Abstract

The second edition of the textbook "Economics and Computation: An Introduction to Algorithmic Game Theory, Computational Social Choice, and Fair Division" (edited and coauthored by Jörg Rothe) has just been published by Springer.

It explores the interrelations between algorithmic game theory, computational social choice, and fair division, three vibrant areas at the interface between economics, collective decision-making, and computer science. In all these parts, much weight is given to the algorithmic and complexity-theoretic aspects of problems arising in these areas, and the interconnections between the three parts are of central interest. In this talk, an overview of the content of this textbook will be given and some gems from these areas will be presented.

Khaled Belahcène

Bio

Khaled Belahcene is junior professor at the MICS lab at CentraleSupélec - Paris Saclay University in France. His research interests encompass decisions, preferences, and explanations. Khaled has been co-chair of the DA2PL 2022 workshop (From Multiple Criteria Decisions Aid to Preference Learning) and is co-editor of an upcoming special issue on preference handling of the Annals of Mathematics and Artificial Intelligence.

Title

Lincs, a Python library for non-compensatory sorting

Abstract

In the field of multiple criteria decision theory, compensation occurs when, no matter how poor an alternative performs on some criterion, it can obtain a favorable decision thanks to a high performance according to another dimension. This behaviour is not always suitable, and models have been proposed to shut it down when the decision to be taken is about positioning alternatives into categories ordered by desirability ("multicriteria sorting"). These models offer concise and easy-to-interpret rules upon which recommender systems or collaborative decision-making systems might be built. As we shall see, these models stemming from decision theory are very close to the models of the ELECTRE family used in MCDA and to the scoring systems popular in e.g. healthcare. Yet, learning the parameters of a non-compensatory sorting system from data is computationally hard. The literature offers three ways to address this challenging problem, either via mixed integer linear programming, maximization of boolean satisfiability, or evolutionary heuristics, which make it possible to process real-world data and thus address real-world problems. We introduce Lincs, a python library providing an efficient implementation of these state-of-the-art techniques through a simple API.