Inspired by the fact that many combinatorial optimization problems arising
in practice are NP-hard, the design of efficient approximation algorithms h
as been a major research topic for the last years. Since we can not expect
to solve any NP-hard problem in polynomial time, it is meaningful to compro
mise optimality of a solution and settle for a "sufficiently good" solution
that can be computed efficiently in polynomial time.