作者: Per Kristian Lehre , Dogan Corus , Mojgan Pourhassan , Frank Neumann
DOI:
关键词:
摘要: Bi-level optimisation problems have gained increasing interest in the field of combinatorial recent years. With this paper, we start runtime analysis evolutionary algorithms for bi-level problems. We examine two NP-hard problems, generalised minimum spanning tree problem (GMST), and travelling salesman (GTSP) context parameterised complexity. For problem, analyse approaches presented by Hu Raidl (2012) with respect to number clusters that distinguish each other chosen representation possible solutions. Our results show a (1+1) EA working nodes is not fixed-parameter algorithm whereas global structure enables solve time. present hard instances approach are highly complementary proving they other's very efficiently. instance. problem.