Quantitative measure of nonconvexity for black-box continuous functions

Metaheuristic algorithms usually aim to solve nonconvex optimization problems in black-box and high-dimensional scenarios. Characterizing and understanding the properties of nonconvex problems is therefore important for effectively analyzing metaheuristic algorithms and their development, improvement and selection for problem solving. This paper establishes a novel analysis framework called nonconvex ratio analysis, which can characterize nonconvex continuous functions by measuring the degree of nonconvexity of a problem. This analysis uses two quantitative measures: the nonconvex ratio for global characterization and the local nonconvex ratio for detailed characterization. Midpoint convexity and Monte Carlo integral are important methods for constructing the measures. Furthermore, as a practical feature, we suggest a rapid characterization measure that uses the local nonconvex ratio and can characterize certain black-box high-dimensional functions using a much smaller sample. Throughout this paper, the effectiveness of the proposed measures is confirmed by numerical experiments using the COCO function set.

[1]  Bernd Bischl,et al.  Exploratory landscape analysis , 2011, GECCO '11.

[2]  L. Darrell Whitley,et al.  The dispersion metric and the CMA evolution strategy , 2006, GECCO.

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

[4]  Kate Smith-Miles,et al.  Measuring algorithm footprints in instance space , 2012, 2012 IEEE Congress on Evolutionary Computation.

[5]  El-Ghazali Talbi,et al.  Metaheuristics - From Design to Implementation , 2009 .

[6]  David H. Wolpert,et al.  No free lunch theorems for optimization , 1997, IEEE Trans. Evol. Comput..

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

[8]  Anne Auger,et al.  Comparing results of 31 algorithms from the black-box optimization benchmarking BBOB-2009 , 2010, GECCO '10.

[9]  Christian L. Müller,et al.  Global Characterization of the CEC 2005 Fitness Landscapes Using Fitness-Distance Analysis , 2011, EvoApplications.

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

[11]  Harald Niederreiter,et al.  Random number generation and Quasi-Monte Carlo methods , 1992, CBMS-NSF regional conference series in applied mathematics.

[12]  Rémi Bardenet,et al.  Monte Carlo Methods , 2013, Encyclopedia of Social Network Analysis and Mining. 2nd Ed..

[13]  Constantin P. Niculescu,et al.  Convex Functions and Their Applications: A Contemporary Approach , 2005 .

[14]  Jean-Pierre Aubin,et al.  Estimates of the Duality Gap in Nonconvex Optimization , 1976, Math. Oper. Res..

[15]  Kate Smith-Miles,et al.  Cross-disciplinary perspectives on meta-learning for algorithm selection , 2009, CSUR.

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

[17]  John R. Rice,et al.  The Algorithm Selection Problem , 1976, Adv. Comput..

[18]  B. Everitt.  The Cambridge Dictionary of Statistics , 1998 .

[19]  Hiroshi Konno,et al.  On the degree and separability of nonconvexity and applications to optimization problems , 1997, Math. Program..

[20]  S. Vavasis.  COMPLEXITY ISSUES IN GLOBAL OPTIMIZATION: A SURVEY , 1995 .

[21]  Michael Affenzeller,et al.  A Comprehensive Survey on Fitness Landscape Analysis , 2012, Recent Advances in Intelligent Engineering Systems.

[22]  Saman K. Halgamuge,et al.  Exploratory Landscape Analysis of Continuous Space Optimization Problems Using Information Content , 2015, IEEE Transactions on Evolutionary Computation.

[23]  Kazuo Murota,et al.  Discrete convex analysis , 1998, Math. Program..