Peter Anderson


A popular application of genetic algorithms (GAs) is to attempt to generate good, rapid, approximate solutions to NP-complete or NP-hard problems. Previously, in [1], and [4], we introduced a hybrid algorithm combining a GA with simple greedy algorithms applied to the N-Queens problem and to sports tournament scheduling. The greedy algorithm makes locally optimal assignments (to Queens or matches) in some order. We treat that ordering as the sought after goal, and thus work with a population whose individuals are permutations. The subject of the present paper is the problem of graph coloring. We focus the present paper on the single benchmark problem of coloring a three-colorable graph that was constructed as a subgraph of the complete 3-partite graph Kp,q,r in which each edge exists with probability 0.1. (We have applied our method successfully to several other categories of graphs, but present space limitations dictate presenting the results for this special case.) (Refer to PDF file for exact formulas)

Date of creation, presentation, or exhibit



Information Science Innovations (2001) Presented at the ICSC/NAISO Conference, Information Science Innovations (ISI 2001). Held in Dubai, UAE: March 2001. Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works in February 2014.

Document Type

Conference Proceeding

Department, Program, or Center

Chester F. Carlson Center for Imaging Science (COS)


RIT – Main Campus