Collins Conference Room
Seminar
  US Mountain Time

Our campus is closed to the public for this event.

Brian Hayes (Writer, American Scientist)

Abstract.  In the 1960s and 70s Thomas C. Schelling, an economist then at Harvard, developed a simple agent-based simulation in which a mixed population of two races spontaneously breaks down into segregated enclaves, even though no member of the population seeks that outcome. Many variations of Schelling's model have been explored by computational methods, but until recently not much could be said with mathematical certainty. Last year Brandt, Immorlica, Kamath and Kleinberg examined a specific one-dimensional version of the model and produced some interesting analytic results; in particular they showed that the Schelling segregation process is self-limiting, in the sense that racial enclaves do not grow to their maximum possible size. Later, Barmpalias, Elwes and Lewis-Pye performed a similar analysis of a slightly wider class of one-dimensional models. I will give a brief introduction to the work of these two groups (in which I was not a participant) and then discuss yet another one-dimensional model, namely the one that Schelling first described in 1969. I note that although we now have exact solutions for certain models of racial segregation, we do not have an exact solution for racial segregation.

Purpose: 
Research Collaboration
SFI Host: 
Cris Moore and Stephan Mertens

More SFI Events