Matchings model a host of real-world examples like matching passengers to taxi drivers or matching residents to hospitals. Significant work has gone into finding algorithms that guarantee that the matchings returned will satisfy some requisite properties. We aim to focus on matching settings where the preferences of one group of agents, conflict with those of the other agents. Typically, in such settings, commonly used algorithms tend to favour one group over the other.
We would like to find matchings that strike a balance between these two groups and possibly suggest novel approaches to this end. We aim to corroborate these proposed solutions both theoretically and experimentally. That is, we aim to mathematically prove that our algorithms will be fair and efficient and then implement and verify this as well.
School
Computer Science and Engineering
Research Area
Computational social choice | Algorithmic game theory
- Research environment
- Expected outcomes
- Supervisory team
- Reference material/links
- The algorithmic decision theory (ADT) group has significant experience in computational social choice. The successful candidate will work closely with the supervisors.
- Theoretical results on the existence of fair and efficient matchings
- Implementation and simulations of matching instances
