作者: Joshua Knowles , David Corne
关键词:
摘要: We describe, and make publicly available, two problem instance generators for a multiobjective version of the well-known quadratic assignment (QAP). The allow number parameters to be set, including those controlling epistasis inter-objective correlations. Based on these generators, several initial test suites are provided described. For each we measure some global properties and, smallest ones, observations Pareto optimal sets/fronts. Our purpose in providing tools is facilitate ongoing study structure (combinatorial) optimization, its effects search landscape algorithm performance.