Graph partitioning using genetic algorithms with ODPX

In this paper, we study approximate solutions to the extension of the "maximally balanced connected partition problem", whose corresponding decision problem is known to be /spl Nscr//spl Pscr/-complete. We introduce a genetic algorithm with a new crossover operator, called the "order- and distance-preserving crossover" (ODPX) operator, and we compare the results of our algorithm to a well-known deterministic approximation algorithm.