Paper Info
Title
Fair solutions for some multiagent optimization problems
Abstract
We consider optimization problems in a multiagent setting where a solution is evaluated with a vector. Each coordinate of this vector represents an agent's utility for the solution. Due to the possible conflicts, it is unlikely that one feasible solution is optimal for all agents. Then, a natural aim is to find solutions that maximize the satisfaction of the least satisfied agent, where the satisfaction of an agent is defined as his relative utility, i.e., the ratio between his utility for the given solution and his maximum possible utility. This criterion captures a classical notion of fairness since it focuses on the agent with lowest relative utility. We study worst-case bounds on this ratio: for which ratio a feasible solution is guaranteed to exist, i.e., to what extend can we find a solution that satisfies all agents? How can we build these solutions in polynomial time? For several optimization problems, we give polynomial-time deterministic algorithms which (almost always) achieve the best possible ratio.
Year
DOI
Venue
2013
10.1007/s10458-011-9188-z
Autonomous Agents and Multi-Agent Systems
Keywords
DocType
Volume
Multiagent optimization,Combinatorial optimization,Fairness
Journal
26
Issue
ISSN
Citations 
2
1387-2532
9
PageRank 
References 
Authors
0.51
29
3
Name
Order
Citations
PageRank
Bruno Escoffier143037.32
Laurent Gourvès224130.97
Jérôme Monnot351255.74