Individualbased models

Forward iterations of the sort described in the text are examples of individual-based models (IBMs), which are becoming a common tool in ecology (DeAngelis and Gross 1992, van Winkle et al. 1998, Railsback 2001). Such models are used to simulate the behavior of many individuals, so that population processes emerge from individual interactions. In order to do this, one must have the rules that the individuals follow and a major question is then, from where do such rules come? One possibility, of course, is simply to make them up from one's sense of how the system works. Another is to use deterministic or stochastic dynamic programming, as in here, but use the results to predict general rules of thumb, rather than state and time dependent tables of behavior. Still another is to use the method of genetic algorithms (Haupt and Haupt 1998), in which behavior is the result of genetic predispositions that evolve through simulated genetic processes (Robertson et al. 1998, Huse et al. 1999, McGregor and Roitberg 2000, Giske et al. 2002, Strand et al. 2002). Other means of achieving individual adaptation are described in Belew and Mitchell (1996).

0 0

Post a comment