Through a summer fellowship at Bowdoin College, I developed and designed an algorithm to schedule tour guides for the admissions office (see poster). In my own time, I experimented with applying AI techniques to improve the algorithm. I have generalized the scheduling algorithm and am currently working on a web application
As part of a final project in Nature Inspired Computation, my group and I explored the use of Ant Colony Optimization to build efficient neural network structures (minimizing the size of the network without compromising network performance).
During the Spring 2015 semester at Bowdoin College, I taught myself and a professor MPI (Message Passing Interface) standards, leveraging the power of a high performance computing grid, in order to execute complex computational geometry programs. My work resulted in significant speed increases.