Re: measuring multiagent system diversity

[Marcus G. Daniels]

JK> Is this a generic measure of Diversity?

Yeah, my read is that you:

 1) construct a matrix of difference scores between each pairs of agents

 2) identify overlapping clusters using a cut-off distance
    accumulating a set of these cluster groups by using a range of cut-offs,
    i.e. different magnifications of a microscope.

 3) compute the entropy of each grouping, and then integrate entropy across
    groups. 

It looks easy to implement, although it seems like one danger could be
flakey distance metrics.


                  ==================================
   Swarm-Modelling is for discussion of Simulation and Modelling techniques
   esp. using Swarm.  For list administration needs (esp. [un]subscribing),
   please send a message to <address@hidden> with "help" in the
   body of the message.
                  ==================================