From the time of my last entry until the sixteenth of May 2011, I ground out the proof of concept code for both of my research projects, thus bringing an end to my last semester as a grad student. Happily, both of my theses proved to be true. First, the transitive subjective trust network that I designed as a graduate student works exceptionally well. In a prisoner’s dilemma style game consisting of 100 agents using mixed collaborative and competitive strategies, agents using the transitive trust network were able to isolate and cripple the earning power of the smaller competitive / aggressive agent population. The control group of agents relying only on individual memory were not able to adequately defend themselves from the aggressive players. I expected that this would be the case, but it was remarkable to see it writ large in the experimental results.
Second, the Pareto-distribution network topology for the particle swarm outperformed both the wheel and the hub networks in multiple trials on the 2D Rastrigin problem space. Needless to say, I couldn’t be more excited about the results of the research.
Now that the semester is over, and I’ve had a few weeks to graduate, recuperate, and reflect on the research, I am starting to get excited about what comes next. I’m moving to Ann Arbor, Michigan where my wife will begin her own graduate student career. What lies in store for me there?
Can I turn this research into a viable startup? Is there a more interesting problem I could be working on? We shall see.