Maximum Satisfiability: Anatomy of the Fitness Landscape for a Hard Combinatorial Optimization Problem

The fitness landscape of MAX-3-SAT is investigated for random instances above the satisfiability phase transition. This paper includes a scaling analysis of the time to reach a local optimum, the number of local optima, the expected probability of reaching a local optimum as a function of its fitness, the expected fitness found by local search and the best fitness, the probability of reaching a global optimum, the size and relative positions of the global optima, the mean distance between the local and global optima, the expected fitness as a function of the Hamming distance from an optimum and their basins of attraction. These analyses show why the problem becomes hard for local search algorithms as the system size increases. The paper also shows how a recently proposed algorithm can exploit long-range correlations in the fitness landscape to improve on the state-of-the-art heuristic algorithms.

[1]  Florent Krzakala,et al.  A Landscape Analysis of Constraint Satisfaction Problems , 2007, Physical review. E, Statistical, nonlinear, and soft matter physics.

[2]  Peter F. Stadler,et al.  Towards a theory of landscapes , 1995 .

[3]  Hector J. Levesque,et al.  A New Method for Solving Hard Satisfiability Problems , 1992, AAAI.

[4]  Anton V. Eremeev,et al.  Statistical analysis of local search landscapes , 2004, J. Oper. Res. Soc..

[5]  Lee Altenberg,et al.  Fitness Distance Correlation Analysis: An Instructive Counterexample , 1997, ICGA.

[6]  Josselin Garnier,et al.  Efficiency of Local Search with Multiple Local Optima , 2001, SIAM J. Discret. Math..

[7]  Thierry Mora,et al.  Clustering of solutions in the random satisfiability problem , 2005, Physical review letters.

[8]  Hector J. Levesque,et al.  Hard and Easy Distributions of SAT Problems , 1992, AAAI.

[9]  Andrew B. Kahng,et al.  A new adaptive multi-start technique for combinatorial global optimizations , 1994, Oper. Res. Lett..

[10]  Adam Prügel-Bennett,et al.  Symmetry breaking in population-based optimization , 2004, IEEE Transactions on Evolutionary Computation.

[11]  Colin R. Reeves,et al.  Estimating the Number of Solutions for SAT Problems , 2004, PPSN.

[12]  Andrea Montanari,et al.  Instability of one-step replica-symmetry-broken phase in satisfiability problems , 2003, ArXiv.

[13]  Assaf Naor,et al.  Rigorous location of phase transitions in hard optimization problems , 2005, Nature.

[14]  Bart Selman,et al.  Noise Strategies for Improving Local Search , 1994, AAAI.

[15]  David E. Goldberg,et al.  Hierarchical BOA Solves Ising Spin Glasses and MAXSAT , 2003, GECCO.

[16]  Rémi Monasson,et al.  Determining computational complexity from characteristic ‘phase transitions’ , 1999, Nature.

[17]  L. Darrell Whitley,et al.  Polynomial Time Summary Statistics for a Generalization of MAXSAT , 1999, GECCO.

[18]  M. Mézard,et al.  Random K-satisfiability problem: from an analytic solution to an efficient algorithm. , 2002, Physical review. E, Statistical, nonlinear, and soft matter physics.

[19]  Stephen A. Cook,et al.  Characterizations of Pushdown Machines in Terms of Time-Bounded Computers , 1971, J. ACM.

[20]  Tim Jones.  Evolutionary Algorithms, Fitness Landscapes and Search , 1995 .

[21]  Thomas Stützle,et al.  Iterated Robust Tabu Search for MAX-SAT , 2003, Canadian Conference on AI.

[22]  James M. Crawford,et al.  Experimental Results on the Crossover Point in Random 3-SAT , 1996, Artif. Intell..

[23]  Lov K. Grover.  Local search and the local structure of NP-complete problems , 1992, Oper. Res. Lett..

[24]  Rémi Monasson,et al.  Statistical mechanics methods and phase transitions in optimization problems , 2001, Theor. Comput. Sci..

[25]  Benjamín Barán,et al.  Ant Colony Optimization with Adaptive Fitness Function for Satisfiability Testing , 2007, WoLLIC.

[26]  David S. Johnson,et al.  Computers and Intractability: A Guide to the Theory of NP-Completeness , 1978 .

[27]  E. Weinberger,et al.  Correlated and uncorrelated fitness landscapes and how to tell the difference , 1990, Biological Cybernetics.

[28]  Andrew M. Sutton,et al.  A Theoretical Analysis of the k-Satisfiability Search Space , 2009, SLS.

[29]  P. Stadler.  Landscapes and their correlation functions , 1996 .

[30]  Kathleen Steinhöfel,et al.  Analysis of Local Search Landscapes for k-SAT Instances , 2010, Math. Comput. Sci..

[31]  Adam Prügel-Bennett,et al.  Learning the Large-Scale Structure of the MAX-SAT Landscape Using Populations , 2010, IEEE Transactions on Evolutionary Computation.

[32]  Ashraf M. Abdelbar,et al.  Swarm optimization with instinct-driven particles , 2003, The 2003 Congress on Evolutionary Computation, 2003. CEC '03..

[33]  L. Darrell Whitley,et al.  Genetic Algorithm Behavior in the MAXSAT Domain , 1998, PPSN.

[34]  Bernd Freisleben,et al.  Fitness landscape analysis and memetic algorithms for the quadratic assignment problem , 2000, IEEE Trans. Evol. Comput..