作者: Ilias Diakonikolas , Mihalis Yannakakis
DOI: 10.1137/080724514
关键词: Multi-objective optimization 、 Mathematical optimization 、 Spanning tree 、 Combinatorics 、 Mathematics 、 Upper and lower bounds 、 Time complexity 、 Pareto distribution 、 Matching (graph theory) 、 Pareto principle 、 Shortest path problem
摘要: We investigate the problem of computing a minimum set solutions that approximates within specified accuracy $\epsilon$ Pareto curve multiobjective optimization problem. show for broad class biobjective problems (containing many important widely studied such as shortest paths, spanning tree, matching, and others), we can compute in polynomial time an $\epsilon$-Pareto contains at most twice set. Furthermore factor 2 is tight these problems; i.e., it NP-hard to do better. present upper lower bounds three or more objectives, well dual number $k$ which provide good approximation curve.