This lecture briefly re-introduces the Pareto-efficient set and Pareto frontier and then describes different early GA-inspired approaches to find the Pareto-efficient set among a set of conflicting objectives. The key conceptual framework used throughout the lecture is that of a "community" of organisms, selected to live in similar environments, that inhabit different regions of niche space. The entire continuum of niche space is the Pareto frontier, and particular solutions along the frontier (different ways in which objectives can be balanced against each other) are like different species living in the community of optimizers that balance objectives differently.
This lecture also includes some comments about population size and its effect on evolution as well as how phenotypic variance can enhance evolvability.
Whiteboard notes for this lecture can be found at:
No comments:
Post a Comment