The primary assignment for EECS 594, Intro to Adaptive Systems, is to present a research proposal that uses adaptive systems to investigate an issue of interest in your field, and critique the proposals of the other students. Dr. Holland says, wisely, that this is good preparation for academia, where much time is spent in reviewing and preparing proposals.
I fretted over what to propose for awhile. And I remembered one of my “favorite” problem sets from Lada Adamic’s SI 614 networks class last winter. So here is my topic for the research proposal:
In a network where each node knows only a small amount of information about its neighbors, effective techniques for geographic search vary by the network’s graph characteristics, such as the distribution of long- and short-range links. An adaptive geographical search algorithm for networks could develop more appropriate strategies for finding shorter paths through graphs with different properties than a simpler all-purpose algorithm would find. Such an algorithm would need to be able to predict the values of its available next moves based on the local information that each node can provide about its neighbors, adjust values, “remember” its path of graph traversal, and develop backtracking strategies.
I actually have some ideas about how to go about this already, which is good because the first proposal presentations start in just over a month. It would greatly behoove me to complete that particular project early, since things like 622 and my thesis are only going to become more intense and demanding as the semester progresses.