Tunably Rugged Landscapes With Known Maximum and Minimum

We propose NM landscapes as a new class of tunably rugged benchmark problems. NM landscapes are well defined on alphabets of any arity, including both discrete and real-valued alphabets, include epistasis in a natural and transparent manner, are proven to have known value and location of the global maximum and, with some additional constraints, are proven to also have a known global minimum. Empirical studies are used to illustrate that, when coefficients are selected from a recommended distribution, the ruggedness of NM landscapes is smoothly tunable and correlates with several measures of search difficulty. We discuss why these properties make NM landscapes preferable to both NK landscapes and Walsh polynomials as benchmark landscape models with tunable epistasis.

[1]  John H. Holland,et al.  Distributed genetic algorithms for function optimization , 1989 .

[2]  S. Kauffman,et al.  Searching the Clinical Fitness Landscape , 2012, PloS one.

[3]  E. D. Weinberger,et al.  The NK model of rugged fitness landscapes and its application to maturation of the immune response. , 1989, Journal of theoretical biology.

[4]  Jason H. Moore,et al.  The Ubiquitous Nature of Epistasis in Determining Susceptibility to Common Human Diseases , 2003, Human Heredity.

[5]  Colin R. Reeves,et al.  Epistasis in Genetic Algorithms: An Experimental Design Perspective , 1995, ICGA.

[6]  Isabelle Boutron,et al.  Single-Center Trials Show Larger Treatment Effects Than Multicenter Trials: Evidence From a Meta-epidemiologic Study , 2011, Annals of Internal Medicine.

[7]  Inman Harvey,et al.  Unconstrain the Population: The Benefits of Horizontal Gene Transfer in Genetic Algorithms , 2013 .

[8]  Joshua D. Knowles,et al.  Analysis of a complete DNA–protein affinity landscape , 2010, Journal of The Royal Society Interface.

[9]  Jeffrey S. Buzas,et al.  An Analysis of $N\!K$ Landscapes: Interaction Structure, Statistical Properties, and Expected Number of Local Optima , 2014, IEEE Transactions on Evolutionary Computation.

[10]  Colin R. Reeves,et al.  Experiments with Tuneable Fitness Landscapes , 2000, PPSN.

[11]  Andries P. Engelbrecht,et al.  Recent Advances in the Theory and Application of Fitness Landscapes , 2013 .

[12]  C. Reeves,et al.  Properties of fitness functions and search landscapes , 2001 .

[13]  Chuan-Kang Ting,et al.  Effect of model complexity for estimation of distribution algorithm in NK landscapes , 2013, 2013 IEEE Symposium on Foundations of Computational Intelligence (FOCI).

[14]  Melanie Mitchell,et al.  The Performance of Genetic Algorithms on Walsh Polynomials: Some Anomalous Results and Their Explanation , 1991, ICGA.

[15]  Julian Togelius,et al.  Geometric differential evolution , 2009, GECCO '09.

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

[17]  Margaret J. Robertson,et al.  Design and Analysis of Experiments , 2006, Handbook of statistics.

[18]  Colin R. Reeves,et al.  An Experimental Design Perspective on Genetic Algorithms , 1994, FOGA.

[19]  Margaret J. Eppstein,et al.  Very large scale ReliefF for genome-wide association analysis , 2008, 2008 IEEE Symposium on Computational Intelligence in Bioinformatics and Computational Biology.

[20]  Terry Jones,et al.  Fitness Distance Correlation as a Measure of Problem Difficulty for Genetic Algorithms , 1995, ICGA.

[21]  Martin Pelikan.  Analysis of estimation of distribution algorithms and genetic algorithms on NK landscapes , 2008, GECCO '08.

[22]  Richard J. Duro,et al.  Real-Valued Multimodal Fitness Landscape Characterization for Evolution , 2010, ICONIP.

[23]  Dirk Thierens,et al.  Hierarchical problem solving with the linkage tree genetic algorithm , 2013, GECCO '13.

[24]  David Millán-Ruiz,et al.  Matching island topologies to problem structure in parallel evolutionary algorithms , 2013, Soft Computing.

[25]  Stuart A. Kauffman,et al.  The origins of order , 1993 .

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

[27]  Yong Gao,et al.  An Analysis of Phase Transition in NK Landscapes , 2002, J. Artif. Intell. Res..

[28]  Jason H. Moore,et al.  The application of michigan-style learning classifiersystems to address genetic heterogeneity and epistasisin association studies , 2010, GECCO '10.

[29]  Alden H. Wright.  The Computational Complexity of NK Fitness Functions , 2017 .

[30]  Julian Francis Miller,et al.  Information Characteristics and the Structure of Landscapes , 2000, Evolutionary Computation.

[31]  Margaret J. Eppstein,et al.  Team Learning for Healthcare Quality Improvement , 2013, IEEE Access.

[32]  Xiaodong Li,et al.  A framework for generating tunable test functions for multimodal optimization , 2011, Soft Comput..

[33]  Wim Hordijk,et al.  Correlation analysis of coupled fitness landscapes , 2005, Complex..

[34]  Paul Hines,et al.  A “Random Chemistry” Algorithm for Identifying Collections of Multiple Contingencies That Initiate Cascading Failure , 2012, IEEE Transactions on Power Systems.

[35]  Ilaria Giannoccaro,et al.  Complex Systems Methodologies for Behavioural Research in Operations Management: NK Fitness Landscape , 2013 .