Quasi-bistability of walk-based landscape measures in stochastic fitness landscapes

Exploratory landscape analysis is a useful method for algorithm selection, parametrization and creating an understanding of how a heuristic optimization algorithm performs on a problem and why. A prominent family of fitness landscape analysis measures are based on random walks through the search space. However, most of these features were only introduced on deterministic fitness functions and it is unclear, under which conditions walk-based landscape features are applicable to noisy optimization problems. In this paper, we empirically analyze the effects of noise in the fitness function on these measures and identify two dominant regimes, where either the underlying problem or the noise are described. Additionally, we observe how step sizes and walk lengths of random walks influence this behavior.

[1]  Michele Vendruscolo,et al.  Statistical Properties of Neutral Evolution , 2002, Journal of Molecular Evolution.

[2]  Andries Petrus Engelbrecht,et al.  A survey of techniques for characterising fitness landscapes and some possible ways forward , 2013, Inf. Sci..

[3]  Victor Picheny,et al.  Noisy kriging-based optimization methods: A unified implementation within the DiceOptim package , 2014, Comput. Stat. Data Anal..

[4]  D. Ackley.  A connectionist machine for genetic hillclimbing , 1987 .

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

[6]  Kun Li,et al.  A novel optimization hardness indicator based on the relationship between optimization hardness and frequency features of real-parameter problems , 2015, Soft Comput..

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

[8]  Vincenzo Cutello,et al.  Parallel Problem Solving from Nature - PPSN XII , 2012, Lecture Notes in Computer Science.

[9]  Mohammad Kazem Sheikh-El-Eslami,et al.  The design of a risk-hedging tool for virtual power plants via robust optimization approach , 2015 .

[10]  M. Affenzeller,et al.  Offspring Selection: A New Self-Adaptive Selection Scheme for Genetic Algorithms , 2005 .

[11]  Galina Merkuryeva,et al.  Simulation-based fitness landscape analysis and optimisation of complex problems , 2015 .

[12]  Andries Petrus Engelbrecht,et al.  Ruggedness, funnels and gradients in fitness landscapes and the effect on PSO performance , 2013, 2013 IEEE Congress on Evolutionary Computation.

[13]  Yaochu Jin,et al.  A comprehensive survey of fitness approximation in evolutionary computation , 2005, Soft Comput..

[14]  Pascal Kerschke,et al.  Comprehensive Feature-Based Landscape Analysis of Continuous and Constrained Optimization Problems Using the R-Package Flacco , 2017, Studies in Classification, Data Analysis, and Knowledge Organization.

[15]  Yuri Malitsky,et al.  Features for Exploiting Black-Box Optimization Problem Structure , 2013, LION.

[16]  Vesselin K. Vassilev,et al.  Smoothness, ruggedness and neutrality of fitness landscapes: from theory to application , 2003 .

[17]  Sébastien Vérel,et al.  A study of NK landscapes' basins and local optima networks , 2008, GECCO '08.

[18]  Catherine C. McGeoch,et al.  Quantum Annealing amid Local Ruggedness and Global Frustration , 2017, Journal of the Physical Society of Japan.

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

[20]  H. S. Heaps,et al.  The effect of noise on process optimization , 1965 .

[21]  Sébastien Vérel,et al.  ParadisEO-MO: from fitness landscape analysis to efficient local search algorithms , 2013, Journal of Heuristics.

[22]  Pierre Ladevèze,et al.  Error Estimate Procedure in the Finite Element Method and Applications , 1983 .

[23]  Michael Affenzeller,et al.  Optimization knowledge base: an open database for algorithm and problem characteristics and optimization results , 2012, GECCO '12.

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

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

[26]  Michael Affenzeller,et al.  Analysis of uncertainty in engineering design optimization problems , 2016 .

[27]  Anne Auger,et al.  Real-Parameter Black-Box Optimization Benchmarking 2009: Noiseless Functions Definitions , 2009 .

[28]  Marcus Gallagher,et al.  Length Scale for Characterising Continuous Optimization Problems , 2012, PPSN.

[29]  Jürgen Branke,et al.  Evolutionary optimization in uncertain environments-a survey , 2005, IEEE Transactions on Evolutionary Computation.

[30]  Kenneth A. De Jong,et al.  Landscape Features for Computationally Expensive Evaluation Functions: Revisiting the Problem of Noise , 2016, PPSN.

[31]  Jonathan E. Fieldsend,et al.  The Rolling Tide Evolutionary Algorithm: A Multiobjective Optimizer for Noisy Optimization Problems , 2015, IEEE Transactions on Evolutionary Computation.

[32]  Heike Trautmann,et al.  Low-Budget Exploratory Landscape Analysis on Multiple Peaks Models , 2016, GECCO.

[33]  Hans-Georg Beyer,et al.  Local Performance of the (μ/μ, μ)-ES in a Noisy Environment , 2000, FOGA.

[34]  Sébastien Vérel,et al.  Local Optima Networks: A New Model of Combinatorial Fitness Landscapes , 2014, ArXiv.

[35]  Galina Merkuryeva,et al.  Benchmark Fitness Landscape Analysis , 2020 .

[36]  Olivier Teytaud,et al.  Bandit-Based Estimation of Distribution Algorithms for Noisy Optimization: Rigorous Runtime Analysis , 2010, LION.