This page is for my master's thesis project, which I plan on completing at the end of Spring Quarter 2026. I started it Spring Quarter 2025 with Professor Daniel Frishberg, and I'll update this page as I make more progress.

The overall goal of the project is to test different Travelling Salesperson algorithms against each other on different kinds of point sets to compare real world performance of these algorithms against their theoretical runtimes. I have long been interested in the TSP problem, and got especially interested in it when I took CSC 549 with Professor Frishberg and did a project on Christofides' Algorithm, which is linked here.

As of now (June 2025), I've implemented the Multifragment Heuristic in C++, and use a linear search of all neighbors to find the nearest neighbor. Professor Frishberg and I also did a lot of theoretical setup for understanding nearest-neighbor chain method for finding neighbors, with the end goal of implementing a soft nearest neighbor chain implementation of the multifragment heuristic. I'll probably also implement Christofides' Algorithm at some point, and then our goal for right now is to try all the different algorithms on real world point sets, probably from TSPLIB.