Michael McKay

Email: 2093581m@student.gla.ac.uk

Research title: Algorithms for matching problems involving preferences

Research Summary

I am a postgraduate research student supervised by Prof. David Manlove. In general, my research is related to algorithms and complexity, matching under preferences, and game theory. I am also interested in graph theory, computational social choice, and wider aspects of algorithms and complexity.

In particular, my thesis work relates to coalition formation problems in which a constraint exists on the sizes of possible coalitions. Many such problems have been studied in the theory of matching under preferences but there is also a very close relation to hedonic games.

To be concrete, such problems could be the assignment of students in a classroom to equal-sized lab groups. With the goal of doing this fairly, we might model the students' preferences over possible groups. One way to do this could be asking each student to list some of their friends. Next we might ask questions such as: what would a fair and optimal assignment look like? Could we produce an assignment in which most students work with at least a few of their friends? Is such an assignment bound to exist? Could an optimal assignment be found using an algorithm? 

Although such problems are encountered in the classroom, other applications relate to robotics or other scenarios of human interaction. In any case, such research helps us understand more about game theory, fair assignment and the theory of algorithms.

My personal website is https://mckay2.github.io/.

Publications

Teaching

I have been a demonstrator and tutor in undergraduate computing science labs at the University of Glasgow. I have tutored:

2020

COMPSCI1023 Web Application Systems (Graduate Apprenticeship Course)

 

COMPSCI2003 Algorithmic Foundations 2

2019

COMPSCI2021 Web Application Development 2

2018

COMPSCI1006 Computing Science 1F - Computing Fundamentals