Gray Box Optimization for Mk Landscapes (NK Landscapes and MAX-kSAT)

This article investigates Gray Box Optimization for pseudo-Boolean optimization problems composed of M subfunctions, where each subfunction accepts at most k variables. We will refer to these as Mk Landscapes. In Gray Box Optimization, the optimizer is given access to the set of M subfunctions. We prove Gray Box Optimization can efficiently compute hyperplane averages to solve non-deceptive problems in time. Bounded separable problems are also solved in time. As a result, Gray Box Optimization is able to solve many commonly used problems from the evolutional computation literature in evaluations. We also introduce a more general class of Mk Landscapes that can be solved using dynamic programming and discuss properties of these functions. For certain type of problems Gray Box Optimization makes it possible to enumerate all local optima faster than brute force methods. We also provide evidence that randomly generated test problems are far less structured than those found in real-world problems.

[1]  Jeremy Frank,et al.  When Gravity Fails: Local Search Topology , 1997, J. Artif. Intell. Res..

[2]  Alden H. Wright,et al.  The computational complexity of N-K fitness functions , 2000, IEEE Trans. Evol. Comput..

[3]  Ton Kloks.  Treewidth, Computations and Approximations , 1994, Lecture Notes in Computer Science.

[4]  Endre Boros,et al.  Pseudo-Boolean optimization , 2002, Discret. Appl. Math..

[5]  William F. Punch,et al.  Hyperplane Elimination for Quickly Enumerating Local Optima , 2016, EvoCOP.

[6]  Thomas Stützle,et al.  Stochastic Local Search: Foundations & Applications , 2004 .

[7]  Sébastien Vérel,et al.  First-Improvement vs. Best-Improvement Local Optima Networks of NK Landscapes , 2010, PPSN.

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

[9]  F. Guerra.  Spin Glasses , 2005, cond-mat/0507581.

[10]  Bart Selman,et al.  Incomplete Algorithms , 2021, Handbook of Satisfiability.

[11]  Peter C. Cheeseman,et al.  Where the Really Hard Problems Are , 1991, IJCAI.

[12]  John H. Holland,et al.  Building Blocks, Cohort Genetic Algorithms, and Hyperplane-Defined Functions , 2000, Evolutionary Computation.

[14]  Doug Hains,et al.  Greedy or Not? Best Improving versus First Improving Stochastic Local Search for MAXSAT , 2013, AAAI.

[15]  L. D. Whitley,et al.  The No Free Lunch and problem description length , 2001 .

[16]  E. Weinberger.  NP Completeness of Kauffman's N-k Model, A Tuneable Rugged Fitness Landscape , 1996 .

[17]  L. Darrell Whitley,et al.  Search, Binary Representations and Counting Optima , 1999 .

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

[19]  Kalyanmoy Deb,et al.  Analyzing Deception in Trap Functions , 1992, FOGA.

[20]  Thomas Jansen,et al.  Analyzing Evolutionary Algorithms: The Computer Science Perspective , 2012 .

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

[22]  Pierre Hansen,et al.  The basic algorithm for pseudo-Boolean programming revisited , 1988, Discret. Appl. Math..

[23]  L. Darrell Whitley,et al.  A Tractable Walsh Analysis of SAT and its Implications for Genetic Algorithms , 1998, AAAI/IAAI.

[24]  Steven E. Hampson,et al.  Large plateaus and plateau search in Boolean Satisfiability problems: When to give up searching and start again , 1993, Cliques, Coloring, and Satisfiability.

[25]  David E. Goldberg,et al.  Genetic Algorithms and Walsh Functions: Part I, A Gentle Introduction , 1989, Complex Syst..

[26]  L. Darrell Whitley,et al.  The Only Challenging Problems Are Deceptive: Global Search by Solving Order-1 Hyperplanes , 1991, ICGA.

[27]  Andrew M. Sutton,et al.  Efficient identification of improving moves in a ball for pseudo-boolean problems , 2014, GECCO.

[28]  Thomas Jansen,et al.  Analyzing Evolutionary Algorithms , 2015, Natural Computing Series.

[29]  L. Darrell Whitley,et al.  Partition Crossover for Pseudo-Boolean Optimization , 2015, FOGA.

[30]  Robert B. Heckendorn.  Embedded Landscapes , 2002, Evolutionary Computation.

[31]  S. Kauffman,et al.  Towards a general theory of adaptive walks on rugged landscapes. , 1987, Journal of theoretical biology.

[32]  John H. Holland,et al.  Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence , 1992 .

[33]  L. Darrell Whitley,et al.  Constant time steepest descent local search with lookahead for NK-landscapes and MAX-kSAT , 2012, GECCO '12.

[34]  L. Darrell Whitley,et al.  Tunnelling Crossover Networks , 2015, GECCO.

[35]  Clifford Stein,et al.  Introduction to Algorithms, 2nd edition. , 2001 .

[36]  Carlos Ansótegui,et al.  The Community Structure of SAT Formulas , 2012, SAT.

[37]  Janet Wiles,et al.  A comparison of neutral landscapes - NK, NKp and NKq , 2002, Proceedings of the 2002 Congress on Evolutionary Computation. CEC'02 (Cat. No.02TH8600).

[38]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[39]  Johan Håstad,et al.  Some optimal inapproximability results , 2001, JACM.

[40]  L. Darrell Whitley,et al.  Mk Landscapes, NK Landscapes, MAX-kSAT: A Proof that the Only Challenging Problems are Deceptive , 2015, GECCO.

[41]  Yong Gao,et al.  On the Treewidth of NK Landscapes , 2003, GECCO.