COMPARATIVE ANALYSIS OF COLORING ALGORITHMS FOR ORDINARY WEIGHTED GRAPH

Abstract

Algorithms of solving the “minimax” weighted graph coloring problem are considered and compared. The mathematical formulation of the problem is presented; the solution optimality criteria are stated. The exact algorithm that always finds the minimum count of colors through Maghout method and then finds the “minimax” option using 3 versions of the critical path method is described. Three fast heuristic algorithms are described: the algorithm that works with the list of vertexes arranged by the local degrees; the algorithm based on the removal of the points and adjacent edges; the algorithm using the vertex saturation rate. All algorithms are considered with examples. To evaluate the algorithm efficiency, a computational experiment on several hundred of randomly generated graphs is set up. The algorithms were compared by the operating speed and the proximity of the result to the "minimax" version of coloring.