An MCMC Based Course to Teaching Assistant Allocation

Abstract

Allotting Teaching Assistants (TAs) to courses is a common task at university centers which typically demands a good amount of human effort. We propose a method to allocate using computer algorithm. The presence of conflicting constraints, posed by requirements which determine tradeoff among them tend to make this problem difficult to solve. This is essentially a matching problem and in this paper has been modeled as a Markov Chain of various intermediate allotments. Later we perform simple Monte-Carlo simulations over a naive bucket filling allotment. This leads us to a globally optimal allotment with a promise of faster convergence.