Fairness in Multiagent Resource Allocation with Dynamic and Partial Observations: Extended Abstract
Abstract
We investigate fairness issues in distributed resource allocation of indivisible goods. More specifically, we study envy-freeness in a setting where the observations of agents only result from encounters with other agents. Agents thus have a partial and uncertain view of the entire allocation, that they maintain throughout the process, and which allows them to have different estimates of their envy. We provide a fully distributed protocol allowing to guarantee termination despite the limited knowledge of agents.
Domains
Artificial Intelligence [cs.AI]
Origin : Files produced by the author(s)