Abstract
This paper combines local search and approximation into “approximation of local optima”, i.e., an attempt to escape hardness of exact local optimization by trying to find solutions which are approximately as good as the worst local optimum. The complexity of several well known optimization problems under this approach is investigated. Our main tool is a special reduction called the “strong PLS-reduction” which preserves cost and local search structure in a very strict sense. Completeness for the class of NP-optimization/local search problems under this reduction allows to deduce that both approximation of global and of local optima cannot be achieved efficiently (unless P = NP resp. P = PLS). We show that the (weighted) problems Min 4-DNF, Max Hopfield, Min/Max 0-1-Programming, Min Independent Dominating Set, and Min Traveling Salesman are all complete under the new reduction.
Moreover the unweighted Min 3-DNF problem is shown to be complete for the class of NP-optimization problems with polynomially bounded cost functions under an approximation preserving reduction. This implies that the logically defined class Min ∑0, the minimization analogue of max SNP, does not capture any (low) approximation degree.
Research supported by DFG-grant Sehn 503/1-1
Preview
Unable to display preview. Download preview PDF.
References
P. Berman, G. Schnitger. On the Complexity of Approximating the Independent Set Problem. Inform. and Comp., vol.96, pp. 77–94, 1992.
M.R. Garey, D.S. Johnson. Computers and Intractability: A Guide to the Theory of NP-Completeness. Freeman, San Francisco, 1979.
M.M. Halldórson. Approximating the minimum maximal independence number. Inf. Proc. Lett., Vol.46, pp. 169–172, 1993.
J. Hertz, A. Krogh, R.G. Palmer. Introduction to the Theory of Neural Computation. Addison-Wesley, 1991.
J.J. Hopfield. Neural Networks and Physical Systems having Emergent Collective Computational Abilities, 1982. Reprinted in Anderson, Rosenfeld: Neurocomputing: Found, of Research, Cambridge, MIT-Press, 1988.
J.J. Hopfield, D.W. Tank. “Neural” Computation of Decisions in Optimization Problems. Biological Cybernetics, vol.52, pp. 141–152, 1985.
D.S. Johnson, C.H. Papadimitriou, M. Yannakakis. How Easy is Local Search? Journ. of Computer and System Sciences, vol.37, pp. 79–100, 1988.
V. Kann. On the Approximability of NP-complete Optimization Problems. Dissertation, NADA Stockholm, 1992.
V. Kann. Polynomially bounded minimization problems which are hard to approximate. Proc. 20th Int. Coll. on Automata, Lang, and Prog., pp. 52–63, 1993, Springer LNCS 700.
S. Khanna, R. Motwani, M. Sudhan, U. Vazirani. On Syntactic versus Computational Views of Approximability. 35th Symp. Found. Comput. Science, pp. 819–830, 1994.
P.G. Kolaitis, M.N. Thakur. Logical definability of NP optimization problems. Inform. and Comp., vol. 115, pp. 321–353, 1994.
M.W. Krentel. Structure in locally optimal solutions. 30th Symp. Found. Comput. Science, pp. 216–221, 1989.
P. Orponen, H. Mannila. On approximation preserving reductions: Complete problems and robust measures. Technical Report C-1987-28, Department of Computer Science, University of Helsinki.
C.H. Papadimitriou, A.A. Schäffer, M. Yannakakis. On the Complexity of Local Search. Proc. 22th ACM Symp. on Theory of Comp., pp. 438–445, 1990.
C.H. Papadimitriou, M. Yannakakis. Optimization, Approximation, and Complexity Classes. Journ. of Computer and System Sciences, vol.43, pp. 425–440, 1991.
A.A. Schäffer, M. Yannakakis. Simple Local Search Problems that are Hard to Solve. SIAM Journal Comput., vol.20, pp. 56–87, 1991.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1996 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Klauck, H. (1996). On the hardness of global and local approximation. In: Karlsson, R., Lingas, A. (eds) Algorithm Theory — SWAT'96. SWAT 1996. Lecture Notes in Computer Science, vol 1097. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-61422-2_123
Download citation
DOI: https://doi.org/10.1007/3-540-61422-2_123
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-61422-7
Online ISBN: 978-3-540-68529-6
eBook Packages: Springer Book Archive