ABSTRACT
This chapter examines how placing an evolutionary algorithm on a lattice affects the overall performance of the algorithm and the connectivity of the lattice. Falling under the general heading of Evolutionary Algorithms, these include the Genetic Algorithm approach and Evolution Strategies schemes. These methods have achieved considerable success in providing speedy pathways toward approaching optimal or near-optimal solutions. Optimization occurs in each of these models through specific processes of selection, recombination, mutation and replacement. Possible applications of this model include cases in which flexible connectivity would benefit optimization. One natural arena in which such a procedure might be found useful is in the domain of parallel computing. Several units, working on a particular problem, might use fitness criteria to determine whether or not it would be best to make or sever a connection. Then optimization might proceed through the best networking of these units.
