For this seminar's project, we implemented a Distributed Graph Processing System (GPS) that used the "Think-Like-A-Vertex" (TLAV) Programming Model and Akka Actors. We recognized a natural correspondence between the TLAV graph processing model and the general actor model for distributed computation and decided Akka would be a solid foundation on which to build a simple but robust GPS, which we believed to be a novel contribution. We achieved the following goals:

• Implement a graph processing system that can be deployed and run in a distributed environment with multiple machines (e.g. AWS EC2), using an actor system to coordinate and synchronize computational nodes;
• Provide users with a simple, extensible API with which to “plug-in" their own graph algorithm implementations.
  • See figure above for an example of the distributed Graph Colouring algorithm implemented in Akka-GPS.

In the beginning of 2020, I participated in an industrial-academic partnership with Kinaxis and Carleton University. My research produced a graphical model of supply chain structures and a heuristic algorithm that detected underutilization in supply networks.
When I took CPSC 547, an Information Visualization graduate course, I extended this work into an interactive graph visualization tool that allows supply chain planners to visualize the structures of their networks, and helps them identify underutilization and anomalies.
· Final Report
· Final Presentation

In November 2019, the Canadian Statistical Sciences Institute hosted its National Case Study Competition. The competition involved a statistical prediction task: given traffic and weather patterns around Metro Vancouver, teams were tasked with the prediction of a probability of delay for BC Ferries sailings.
Our team of Carleton University students developed an ensemble that placed 2^nd in the national competition. In the finals, we presented our methodology and derived feature importances to members of the BC Ferries executive team.
· Poster