Document Type

Honors Project On-Campus Access Only


This paper aims to build connections between representation theory and signal processing. We consider ranked voting data, such as when voters rank n candidates or when users rank n movies. Each ranking is a permutation, and the total count of votes for each permutation can be viewed as a vector on the permutohedron. In order to analyze this data using techniques from graph signal processing, it is helpful to understand the eigenvalues and eigenvectors of the permutohedron and other Cayley graphs of the symmetric group. We use tools from representation theory to reduce the problem of computing n! eigenvalues and vectors to computations in smaller matrices, and we use methods from graph signal processing to find partisan patterns in the voting data.



© Copyright is owned by author of this document