Abstract
Metaheuristics are randomised search algorithms that are effective at finding “good enough” solutions to optimisation problems. However, they present no justification for generated solutions and these solutions are non-trivial to analyse in most cases. We propose that identifying the combinations of variables that strongly influence solution quality, and the nature of this relationship, represents a step towards explaining the choices made by a metaheuristic. Using three benchmark problems, we present an approach to mining this information by using a “surrogate fitness function” within a metaheuristic. For each problem, rankings of the importance of each variable with respect to fitness are determined through sampling of the surrogate model. We show that two of the three surrogate models tested were able to generate variable rankings that agree with our understanding of variable importance rankings within the three common binary benchmark problems trialled.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
Version 3.8.5—https://www.cs.waikato.ac.nz/ml/weka/.
Reference
Adadi, A., Berrada, M.: Peeking inside the black-box: a survey on explainable artificial intelligence (XAI). IEEE Access 6, 52138–52160 (2018). https://doi.org/10.1109/ACCESS.2018.2870052
Brownlee, A.E.I., McCall, J.A.W., Shakya, S.K., Zhang, Q.: Structure learning and optimisation in a markov-network based estimation of distribution algorithm. In: Proceedings of IEEE CEC, pp. 447–454. IEEE Press, Trondheim, Norway (2009)
Brownlee, A.E.I., Regnier-Coudert, O., McCall, J.A.W., Massie, S.: Using a Markov network as a surrogate fitness function in a genetic algorithm. In: Proceedings of IEEE CEC, pp. 4525–4532. Barcelona (2010)
Brownlee, A.E.I., Woodward, J., Swan, J.: Metaheuristic design pattern: surrogate fitness functions. In: MetaDeeP Workshop, Proceedings of GECCO Companion, pp. 1261–1264. ACM Press, Madrid, Spain (2015)
Brownlee, A., McCall, J., Zhang, Q.: Fitness modeling with Markov networks. IEEE Trans. Evolut. Comput. 17(6), 862–879 (2013)
Brownlee, A.E.I., Regnier-Coudert, O., McCall, J.A.W., Massie, S., Stulajter, S.: An application of a GA with Markov network surrogate to feature selection. Int. J. Syst. Sci. 44(11), 2039–2056 (2013)
Brownlee, A.E.I., Wallace, A., Cairns, D.: Mining Markov network surrogates to explain the results of metaheuristic optimisation. In: Proceedings of the SICSA XAI Workshop 2021, CEUR Workshop Proceedings, Robert Gordon University, Aberdeen, UK (2021)
Brownlee, A.E., Weiszer, M., Chen, J., Ravizza, S., Woodward, J.R., Burke, E.K.: A fuzzy approach to addressing uncertainty in Airport Ground Movement optimisation. Trans. Res. Part C Emerg. Technol. 92, 150–175 (2018). https://doi.org/10.1016/j.trc.2018.04.020
Brownlee, A.E., Wright, J.A.: Constrained, mixed-integer and multi-objective optimisation of building designs by NSGA-II with fitness approximation. Appl. Soft Comput. 33, 114–126 (2015)
Chugh, T., Rahat, A., Volz, V., Zaefferer, M.: Towards better integration of surrogate models and optimizers. In: Bartz-Beielstein, T., Filipič, B., Korošec, P., Talbi, E.-G. (eds.) High-Performance Simulation-Based Optimization. SCI, vol. 833, pp. 137–163. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-18764-4_7
Deb, K., Srinivasan, A.: Innovization: Discovery of innovative design principles through multiobjective evolutionary optimization. In: Knowles, J., et al. (eds.) Multiobjective Problem Solving from Nature: From Concepts to Applications, pp. 243–262. Springer, Heidelberg (2008). https://doi.org/10.1007/978-3-540-72964-8_12
Deb, K., Bandaru, S., Greiner, D., Gaspar-Cunha, A., Tutum, C.C.: An integrated approach to automated innovization for discovering useful design principles: case studies from engineering. Appl. Soft Comput. 15, 42–56 (2014). https://doi.org/10.1016/j.asoc.2013.10.011
Gaier, A., Asteroth, A., Mouret, J.B.: Data-efficient exploration, optimization, and modeling of diverse designs through surrogate-assisted illumination. In: Proceedings of the Genetic and Evolutionary Computation Conference, pp. 99–106 (2017)
Hauschild, M., Pelikan, M.: An introduction and survey of estimation of distribution algorithms. Swarm Evol. Comput. 1(3), 111–128 (2011)
Jin, Y.: A comprehensive survey of fitness approximation in evolutionary computation. Soft Comput. 9(1), 3–12 (2005)
Jin, Y.: Surrogate-assisted evolutionary computation: recent advances and future challenges. Swarm Evol. Comput. 1(2), 61–70 (2011)
Larrañaga, P., Lozano, J.A.: Estimation of Distribution Algorithms: A New Tool for Evolutionary Computation. Kluwer, Boston (2002)
Le Bras, P., Robb, D.A., Methven, T.S., Padilla, S., Chantler, M.J.: Improving user confidence in concept maps: Exploring data driven explanations. In: Conference on Human Factors in Computing Systems - Proceedings 2018-April (2018). https://doi.org/10.1145/3173574.3173978
Lehman, J., Clune, J., Misevic, D.: The surprising creativity of digital evolution: A collection of anecdotes from the evolutionary computation and artificial life research communities. Artif. Life 26(2), 274–306 (2020). https://doi.org/10.1162/artl_a_00319
Lozano, J.A., Larrañaga, P., Inza, I., Bengoetxea, E.: Towards a New Evolutionary Computation: Advances on Estimation of Distribution Algorithms (Studies in Fuzziness and Soft Computing). Springer, Heidelberg (2006)
Mouret, J.B., Clune, J.: Illuminating search spaces by mapping elites. arXiv preprint arXiv:1504.04909 (2015)
Ochoa, G., Christie, L.A., Brownlee, A.E., Hoyle, A.: Multi-objective evolutionary design of antibiotic treatments. Artif. Intell. Med. 102, 101759 (2020). https://doi.org/10.1016/j.artmed.2019.101759
Rodriguez Rafael, G.D., Solano Salinas, C.J.: Empirical study of surrogate models for black box optimizations obtained using symbolic regression via genetic programming. In: Proceedings of GECCO Companion, pp. 185–186. ACM (2011)
Sajja, P.S.: Examples and applications on genetic algorithms. In: Illustrated Computational Intelligence. SCI, vol. 931, pp. 155–189. Springer, Singapore (2021). https://doi.org/10.1007/978-981-15-9589-9_5
Urquhart, N., Guckert, M., Powers, S.: Increasing trust in meta-heuristics using MAP-elites. In: Proceedings of GECCO Computation, pp. 1345–1348 (2019)
Wright, J.A., et al.: Multi-objective optimization of cellular fenestration by an evolutionary algorithm. J. Build. Perform. Sim. 7(1), 33–51 (2014)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Wallace, A., Brownlee, A.E.I., Cairns, D. (2021). Towards Explaining Metaheuristic Solution Quality by Data Mining Surrogate Fitness Models for Importance of Variables. In: Bramer, M., Ellis, R. (eds) Artificial Intelligence XXXVIII. SGAI-AI 2021. Lecture Notes in Computer Science(), vol 13101. Springer, Cham. https://doi.org/10.1007/978-3-030-91100-3_5
Download citation
DOI: https://doi.org/10.1007/978-3-030-91100-3_5
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-91099-0
Online ISBN: 978-3-030-91100-3
eBook Packages: Computer ScienceComputer Science (R0)