On Fixed-Parameter Tractability and Approximability of NP Optimization Problems

AbstractFixed-parameter tractability of NP optimization problems is studied by relating it to approximability of the problems. It is shown that an NP optimization problem is fixed-parameter tractable if it admits a fully polynomial-time approximation scheme, or if it belongs to the class MAX SNP or to the class MIN F+Π1. This provides strong evidence that noW[1]-hard NP optimization problems belong to these optimization classes and includes a very large class of approximable optimization problems into the class of fixed-parameter tractable problems. Evidence is also demonstrated to support the current working hypothesis in the theory of fixed-parameter tractability