{"references": ["H. Ishibuchi, H. Tanaka, and H. Okada, Fuzzy neural networks with fuzzy\nweights and fuzzy biases, Proc. of IEEE International Conferences on\nNeural Networks, pp.1650-1655, 1993.", "D.B. Fogel, L.J. Fogel, and V.W. Porto, Evolving neural networks,\nBiological Cybernetics, vol.63, issue 6, pp.487-493, 1990.", "X. Yao, Evolving artificial neural networks, Proc. of the IEEE, vol.87,\nno.9, pp.1423-1447, 1999.", "K.O. Stanley and R. Miikkulainen, Evolving neural networks\nthrough augmenting topologies, Evolutionary Computation, vol.10, no.2,\npp.99-127, 2002.", "D. Floreano, P. Durr, and C. Mattiussi, Neuroevolution: From\narchitectures to learning, Evolutionary Intelligence, vol.1, no.1, pp.47-62,\n2008.", "H. Okada, Genetic algorithm with fuzzy genotype values and its\napplication to neuroevolution, International Journal of Computer,\nInformation Science and Engineering, vol.8, no.1, pp.1-7, 2014.", "H-P. Schwefel, Evolution and Optimum Seeking, Wiley, 1995.", "H-G. Beyer and H-P. Schwefel, Evolution strategies - A comprehensive\nintroduction, Natural Computing, vol.1, no.1, pp.3-52, 2002.", "N. Hansen and A. Ostermeier, Adapting arbitrary normal mutation\ndistributions in evolution strategies: The covariance matrix adaptation,\nProc. of the 1996 IEEE International Conference on Evolutionary\nComputation, pp.312-317, 1996.\n[10] M. Herdy, Evolution strategies with subjective selection, Proc. of the\n4th International Conference on Parallel Problem Solving from Nature,\npp.22-31, 1996.\n[11] J. Knowles and D. Corne, The pareto archived evolution strategy: A\nnew baseline algorithm for pareto multiobjective optimisation, Proc. of\nthe 1999 Congress on Evolutionary Computation, pp.98-105, 1999.\n[12] N. Hansen and A. Ostermeier, Completely derandomized self-adaptation\nin evolution strategies, Evolutionary Computation, vol.9, no.2,\npp.159-195, 2001.\n[13] H.G. Beyer, The Theory of Evolution Strategies, Springer, 2001.\n[14] A. Auger and N. Hansen, A restart CMA evolution strategy with\nincreasing population size, Proc. of the 2005 IEEE Congress on\nEvolutionary Computation, pp.1769-1776, 2005.\n[15] O.M. Shir and T. Back, Niching in evolution strategies, Proc. of the\n2005 Conference on Genetic and Evolutionary Computation, pp.915-916,\n2005.\n[16] N. Hansen, The CMA evolution strategy: A comparing review, Towards\na New Evolutionary Computation, Springer, pp.1769-1776, 2006.\n[17] D.V. Arnold, Weighted multirecombination evolution strategies,\nTheoretical Computer Science - Foundations of Genetic Algorithms,\nvol.361, no. 1, pp.18-37, 2006.\n[18] X. Chen, X. Liu, and Y. Jia, Combining evolution strategy and gradient\ndescent method for discriminative learning of bayesian classifiers, Proc. of\nthe 11th Annual Conference on Genetic and Evolutionary Computation,\npp.507-514, 2009.\n[19] D.V. Arnold and A.S. Castellarin, A novel approach to adaptive isolation\nin evolution strategies, Proc. of the 11th Annual Conference on Genetic\nand Evolutionary Computation, pp.491-498, 2009.\n[20] L. Graening, N. Aulig, and M. Olhofer, Towards directed open-ended\nsearch by a novelty guided evolution strategy, Proc. of the 11th\nInternational Conference on Parallel Problem Solving from Nature: Part\nII, pp.71-80, 2010.\n[21] A. Auger, D. Brockhoff, and N. Hansen, Analyzing the impact of\nmirrored sampling and sequential selection in elitist evolution strategies,\nProc. of the 11th Workshop on Foundations of Genetic Algorithms,\npp.127-138, 2011.\n[22] A. Auger, D. Brockhoff, and N. Hansen, Mirrored sampling in evolution\nstrategies with weighted recombination, Proc. of the 13th Annual\nConference on Genetic and Evolutionary Computation, pp.861-868, 2011.\n[23] I. Loshchilov, M. Schoenauer, and M. Sebag, Self-adaptive\nsurrogate-assisted covariance matrix adaptation evolution strategy,\nProc. of the Fourteenth International Conference on Genetic and\nEvolutionary Computation Conference, pp.321-328, 2012.\n[24] J. Lu, B. Li, and Y. Jin, An evolution strategy assisted by an ensemble of\nlocal gaussian process models. Proc. of the Fifteenth Annual Conference\non Genetic and Evolutionary Computation Conference, pp.447-454, 2013.\n[25] R. Li, M.T.M. Emmerich, J. Eggermont, T. Back, M. Schutz, J. Dijkstra,\nand J.H.C. Reiber, Mixed integer evolution strategies for parameter\noptimization, Evolutionary Computation, vol.21, no.1 pp.29-64, 2013.\n[26] L.A. Zadeh, The concept of a linguistic variable and its application\nto approximate reasoning - I, II, and III, Information Sciences, vol.8,\npp.199-249, pp.301-357, and vol.9, pp.43-80, 1975.\n[27] G. Alefeld and J. Herzberger, Introduction to Interval Computation,\nAcademic Press, 1983.\n[28] H. Okada, J. Tokida, and Y. Fujii, Comparison of evolution strategy,\ngenetic algorithm and their hybrids on evolving autonomous game\ncontroller agents, International Journal of Science and Engineering\nInvestigations, 10612-02, vol.1, no. 6, pp.11-16, 2012."]}

Evolution strategy (ES) is a well-known instance of evolutionary algorithms, and there have been many studies on ES. In this paper, the author proposes an extended ES for solving fuzzy-valued optimization problems. In the proposed ES, genotype values are not real numbers but fuzzy numbers. Evolutionary processes in the ES are extended so that it can handle genotype instances with fuzzy numbers. In this study, the proposed method is experimentally applied to the evolution of neural networks with fuzzy weights and biases. Results reveal that fuzzy neural networks evolved using the proposed ES with fuzzy genotype values can model hidden target fuzzy functions even though no training data are explicitly provided. Next, the proposed method is evaluated in terms of variations in specifying fuzzy numbers as genotype values. One of the mostly adopted fuzzy numbers is a symmetric triangular one that can be specified by its lower and upper bounds (LU) or its center and width (CW). Experimental results revealed that the LU model contributed better to the fuzzy ES than the CW model, which indicates that the LU model should be adopted in future applications of the proposed method.