Abstract
Here we describe quality diversity algorithms, a recent and powerful class of evolutionary algorithms that produces a diverse set of high-performing solutions. The optimization paradigm emphasizes phenotypic niching and egalitarian treatment of quality and diversity. We ground quality diversity in ecology, describe the historical development, and give an intuition and formalization of the algorithms. We present a practical example that we refer to for engineers and laymen readers to understand how and why quality diversity can be used. The main insights from research of quality diversity, performance metrics, and benchmarks are discussed. Finally, the open challenges are presented.
An animal’s behaviour tends to maximize the survival of the genes “for” that behaviour, whether or not those genes happen to be in the body of the particular animal performing it.
- Richard Dawkins [13]
There is a power and utility to regarding the gene as the unit of selection, but equally there is value to seeing the organism as the unit of niche construction.
- Kevin Laland [46]
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Popular literature about computational evolution often use the active tense when talking about evolution, as though it is a driving force rather than an emergent property of life. Even natural selection could be seen as a misnomer, as selection implies an entity acting upon the world, whereas selection seems to be an emergent property of complex interactions in nature.
- 2.
Laland takes Dawkins’ idea one step further and introduces the idea of niche construction. By influencing its environment, a creature can create its own niche. This introduces the idea of evolution being a causally cyclical process. A creature creates a niche by acting on its environment, its genome adapts to the niche, which in itself causes changes in the environment, and so forth [46].
- 3.
As in diversity of qualities. Both quality diversity as well as illumination are used in the field, although one could view the deeper concept to be phenotypic niching.
- 4.
The field of evolutionary algorithms does not always make a distinction between genetic or phenotypic niching. Confusingly, after Deb mentions the biological definition of species and niches, he defines niches as being artificial subpopulations and niching as a method to force population diversity [17]. There is a case to be made to use more rigor in the definition of niches and species. Niches are basins of attraction in the objective function in phenotypic space. Species are phenotypic solutions that tend to fill certain niches. The genomes of species tend to be similar or at least compatible. One could then argue that, especially when the phenotype is indirectly encoded in the genome, niching has to take place based on phenotypic characteristics. Speciation still takes place on a genetic level, to ensure compatibility between genes. Placing speciation on this level is compatible with Dawkins’ understanding of genes as a primary evolutionary unit [12].
- 5.
One can also argue that the behavior of a neural robot controller in a particular maze is the extended phenotype’s embedding in its environment. When a controller was evolved to act in a specific environment, it is not of interest to describe its behavior in other environments.
- 6.
Igel and Toussaint analyzed the effects of neutral encodings on computational evolution and showed they are necessary for self-adaptation while only marginally increasing the number of necessary function evaluations [42].
- 7.
Classes translate to species as defined in [3].
References
Alonso, D., Etienne, R.S., McKane, A.J.: The merits of neutral theory. Trends Ecol. Evol. 21(8), 451–457 (2006)
Arulkumaran, K., Cully, A., Togelius, J.: Alphastar: an evolutionary computation perspective. In: Proceedings of the Genetic and Evolutionary Computation Conference Companion, pp. 314–315 (2019)
Basto-Fernandes, V., Yevseyeva, I., Deutz, A., Emmerich, M.: A survey of diversity oriented optimization: problems, indicators, and algorithms. In: EVOLVE–A Bridge Between Probability, Set Oriented Numerics and Evolutionary Computation, vol. 7, pp. 3–23. Springer (2017)
Basto-Fernandes, V., Yevseyeva, I., Emmerich, M.: A survey of diversity-oriented optimization. EVOLVE 2013-A Bridge Probab. Set Oriented Numer. Evol. Comput. 1(2013), 101–109 (2013)
Cully, A.: Autonomous skill discovery with quality-diversity and unsupervised descriptors. In: Proceedings of the Genetic and Evolutionary Computation Conference, pp. 81–89 (2019)
Cully, A., Clune, J., Tarapore, D., Mouret, J.-B.: Robots that can adapt like animals. Nature 521(7553), 503–507 (2015)
Cully, A., Demiris, Y.: Quality and diversity optimization: a unifying modular framework. In: IEEE Transactions on Evolutionary Computation, pp. 1–15 (2017)
Cully, A., Demiris, Y.: Hierarchical behavioral repertoires with unsupervised descriptors. Presented at the (2018)
Cully, A., Mouret, J.-B.: Learning to walk in every direction. Evol. Comput. 24(1), 59–88 (2013)
Cully, A., Mouret, J.-B.: Evolving a behavioral repertoire for a walking robot. Evol. Comput. 24(1), 59–88 (2016)
Cully, A., Pierre, U., Mouret, J.-B.: Behavioral repertoire learning in robotics. In: Proceeding of the Fifteenth Annual Conference on Genetic and Evolutionary Computation Conference - GECCO ’13, pp. 175–182 (2013)
Dawkins, R.: The Selfish Gene. Oxford University Press, Oxford (1976)
Dawkins, R.: The Extended Phenotype, vol. 8. Oxford University Press, Oxford (1982)
De Jong, K.A.: An analysis of the behavior of a class of genetic adaptive systems (1975)
Deb, K., Agrawal, S., Pratap, A., Meyarivan, T.: A fast elitist non-dominated sorting genetic algorithm for multi-objective optimization: Nsga-ii, pp. 849–858. Springer (2000)
Deb, K., Goldberg, D.E.: An Investigation of Niche and Species Formation in Genetic Function Optimization (1989)
Deb, K., Spears, W.: Speciation methods. Evol. Comput. 2, 93–100 (2010)
Deb, K., Srinivasan, A.: Innovization: innovating design principles through optimization. Presented at the (2006)
Deb, K., Tiwari, S.: Omni-optimizer: a generic evolutionary algorithm for single and multi-objective optimization. Eur. J. Oper. Res. 185(3), 1062–1087 (2008)
Doncieux, S., Coninx, A.: Open-ended evolution with multi-containers QD. In: Proceedings of the Genetic and Evolutionary Computation Conference Companion, pp. 107–108 (2018)
Duarte, M., Gomes, J., Oliveira, S.M., Christensen, A.L.: Evorbc: evolutionary repertoire-based control for robots with arbitrary locomotion complexity. In: Proceedings of the Genetic and Evolutionary Computation Conference 2016, pp. 93–100 (2016)
Ecoffet, A., Huizinga, J., Lehman, J., Stanley, K.O., Clune, J.: Go-explore: a new approach for hard-exploration problems (2019). arXiv:1901.10995
Elman, J.L.: Finding structure in time. Cogn. Sci. 14(1990), 179–211 (1990)
Gaier, A., Asteroth, A., Mouret, J.-B.: Aerodynamic design exploration through surrogate-assisted illumination. In: 18th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference, AIAA AVIATION Forum, (AIAA 2017-3330) (2017)
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. 9–106 (2017)
Gaier, A., Asteroth, A., Mouret, J.-B.: Are quality diversity algorithms better at generating stepping stones than objective-based search? In: Proceedings of the Genetic and Evolutionary Computation Conference Companion, pp. 115–116 (2019)
Goldberg, D.E., Richardson, J.: In: Genetic algorithms with sharing for multimodal function optimization, pp. 41–49. Lawrence Erlbaum, Hillsdale, NJ (1987)
Gomes, J., Lyhne Christensen, A.: Comparing Approaches for Evolving High-level Robot Control based on Behaviour Repertoires. In: 2018 IEEE Congress on Evolutionary Computation (CEC), pp. 1–6 (2018)
Gomes, J., Oliveira, S.M., Christensen, A.L.: An approach to evolve and exploit repertoires of general robot behaviours. In: Swarm and Evolutionary Computation, pp. 265–283 (2018)
Gravina, D., Liapis, A., Yannakakis, G.N.: Surprise search for evolutionary divergence (2017). arXiv:1706.02556
Gravina, D., Liapis, A., Yannakakis, G.N.: Quality diversity through surprise. IEEE Trans. Evol. Comput. PP(c):1 (2018)
Hagg, A., Asteroth, A., Bäck, T.: In: Prototype Discovery Using Quality-diversity, pp. 500–511. Springer, Berlin (2018)
Hagg, A., Asteroth, A., Bäck, T.: Modeling user selection in quality diversity. In: Proceedings of the 2019 on Genetic and Evolutionary Computation Conference - GECCO 2019 (2019)
Hagg, A., Asteroth, A., Thomas, B.: A deep dive into exploring the preference hypervolume. In: ICCC (2020)
Hagg, A., Preuss, M., Asteroth, A., Bäck, T.: An analysis of phenotypic diversity in multi-solution optimization. In: BIOMA 2020 (2020)
Hagg, A., Wilde, D., Asteroth, A., Bäck, T.: Designing air flow with surrogate-assisted phenotypic niching (2020)
Harik, G.R.: Finding multimodal solutions using restricted tournament selection. In: ICGA, pp. 24–31 (1995)
Hart, E., Steyven, A.S.W., Paechter, B.: Evolution of a functionally diverse swarm via a novel decentralised quality-diversity algorithm. In: Proceedings of the Genetic and Evolutionary Computation Conference, pp. 101–108 (2018)
Holland, J.H.: Adaptation in Natural and Artificial Systems. MIT press (1975)
Howard, D., Eiben, A.E., Kennedy, D.F., Mouret, J.-B., Valencia, P., Winkler, D.: Evolving embodied intelligence from materials to machines. Nat. Mach. Intell. 1(1), 12–19 (2019)
Huizinga, J., Clune, J.: Evolving multimodal robot behavior via many stepping stones with the combinatorial multi-objective evolutionary algorithm (2018). arXiv:1807.03392
Igel, C., Toussaint, M.: Neutrality and self-adaptation. Nat. Comput. 2(2), 117–132 (2003)
Jegorova, M., Doncieux, S., Hospedales, T.: Generative adversarial policy networks for behavioural repertoire (2018). arXiv:1811.02945
Kim, S., Doncieux, S.: Learning highly diverse robot throwing movements through quality diversity search. In: Proceedings of the Genetic and Evolutionary Computation Conference Companion, pp. 1177–1178 (2017)
Koos, S., Mouret, J.-B., Doncieux, S.: The transferability approach: crossing the reality gap in evolutionary robotics. In: IEEE Transactions on Evolutionary Computation, pp. 1–25 (2012)
Laland, K.N.: Extending the extended phenotype. Biol. Philos. 19(3), 313–325 (2004)
Lehman, J., Miikkulainen, R.: Enhancing divergent search through extinction events. Presented at the (2015)
Lehman, J., Stanley, K.O.: Exploiting open-endedness to solve problems through the search for novelty. In: Alife, pp. 329–336 (2008)
Lehman, J., Stanley, K.O.: Abandoning objectives: evolution through the search for novelty alone. Evol. Comput. 19(2), 189–222 (2011)
Lehman, J., Stanley, K.O.: Evolving a diversity of virtual creatures through novelty search and local competition. In: Proceedings of the 13th Annual Conference on Genetic and Evolutionary Computation, pp. 211–218 (2011)
Meyerson, E., Miikkulainen, R.: Discovering evolutionary stepping stones through behavior domination. In: Proceedings of the Genetic and Evolutionary Computation Conference, pp. 139–146 (2017)
Mlot, N.J., Tovey, C.A., David, L.H.: Fire ants self-assemble into waterproof rafts to survive floods. Proc. Natl. Acad. Sci. 108(19), 7669–7673 (2011)
Mouret, J.-B.: Novelty-based multiobjectivization. In: New Horizons in Evolutionary Robotics, pp. 139–154. Springer (2011)
Mouret, J.-B., Clune, J.: An algorithm to create phenotype-fitness maps. Proc. Artif. Life Conf. 375(2012), 593–594 (2012)
Mouret, J.-B., Clune, J.: Illuminating search spaces by mapping elites (2015). arXiv:1504.04909
Mouret, J.-B., Doncieux, S.: Encouraging behavioral diversity in evolutionary robotics: an empirical study. Evol. Comput. 20(1), 91–133 (2012)
Nguyen, A., Yosinski, J., Clune, J.: Deep neural networks are easily fooled: high confidence predictions for unrecognizable images. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 427–436 (2015)
Nguyen, A., Yosinski, J., Clune, J.: Innovation engines: automated creativity and improved stochastic optimization via deep learning. In: Proceedings of the 2015 on Genetic and Evolutionary Computation Conference - GECCO ’15, pp. 959–966 (2015)
Nguyen, A., Yosinski, J., Clune, J.: Understanding innovation engines: automated creativity and improved stochastic optimization via deep learning. Evol. Comput. 24(3), 545–572 (2016)
Nordmoen, J., Samuelsen, E., Ellefsen, K.O., Glette, K.: Dynamic Mutation in Map-Elites for Robotic Repertoire Generation, pp. 598–605. MIT Press (2018)
Pétrowski, A.: In: A clearing procedure as a niching method for genetic algorithms, pp. 798–803. IEEE (1996)
Preuss, M.: Niching prospects. In: Proceedings of Bioinspired Optimization Methods and their Applications (BIOMA 2006), pp. 25–34 (2006)
Preuss, M.: Niching the cma-es via nearest-better clustering. In: Proceedings of the 12thAnnual Conference Companion on Genetic and Evolutionary Computation, pp. 1711–1718 (2010)
Preuss, M.: Multimodal Optimization by Means of Evolutionary Algorithms. Springer (2015)
Pugh, J.K., Soros, L.B., Stanley, K.O.: An extended study of quality diversity algorithms. In: Proceedings of the 2016 on Genetic and Evolutionary Computation Conference Companion, pp. 19–20 (2016)
Pugh, J.K., Soros, L.B., Stanley, K.O.: Searching for quality diversity when diversity is unaligned with quality. In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 9921 LNCS:880–889 (2016)
Pugh, J.K., Soros, L.B., Szerlip, P.A., Stanley, K.O.: Confronting the challenge of quality diversity. In: Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation, pp. 967–974 (2015)
Rastrigin, L.A.: Systems of extremal control. Nauka (1974)
Schaaf, L.J., John Odling-Smee, F., Laland, K.N., Feldman, M.W.: Niche Construction: The Neglected Process in Evolution. Princeton University Press, Princeton (2003)
Shir, O.M., Preuss, M., Naujoks, B., Emmerich, M.: Enhancing decision space diversity in evolutionary multiobjective algorithms. In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 5467 LNCS:95–109 (2009)
Sims, K.: Evolution of Virtual Creatures. In: Proceedings of the 21st Annual Conference on Computer Graphics and Interactive Techniques (1994)
Skinner, B.F.: Reinforcement today. Am. Psychol. 13(3), 94 (1958)
Smith, D., Tokarchuk, L., Wiggins, G.: Rapid phenotypic landscape exploration through hierarchical spatial partitioning. In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 9921 LNCS:911–920 (2016)
Stanley, K.O., D’Ambrosio, D.B., Gauci, J.: A hypercube-based encoding for evolving large-scale neural networks. Artif. Life 15(2), 185–212 (2009)
Tarapore, D., Clune, J., Cully, A., Mouret, J.-B.: How do different encodings influence the performance of the map-elites algorithm? In: Proceedings of the Genetic and Evolutionary Computation Conference 2016, pp. 173–180 (2016)
Toffolo, A., Benini, E.: Genetic diversity as an objective in multi-objective evolutionary algorithms. Evol. Comput. 11(2), 151–167 (2003)
Ulrich, T.: Integrating decision space diversity into hypervolume-based multiobjective search categories and subject descriptors. In: GECCO 2010, pp. 455–462 (2010)
Urquhart, N., Hart, E.: Optimisation and illumination of a real-world workforce scheduling and routing application (WSRP) via map-elites. In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 11101 LNCS:488–499 (2018)
Vassiliades, V., Mouret, J.-B.: Discovering the elite hypervolume by leveraging interspecies correlation. In: Proceedings of the Genetic and Evolutionary Computation Conference, pp. 149–156 (2018)
Vassiliades, V., Chatzilygeroudis, K., Mouret, J.-B.: A comparison of illumination algorithms in unbounded spaces. In: Proceedings of the Genetic and Evolutionary Computation Conference Companion, pp. 1578–1581 (2017)
Vassiliades, V., Chatzilygeroudis, K., Mouret, J.-B.: Using centroidal voronoi tessellations to scale up the multidimensional archive of phenotypic elites algorithm. IEEE Trans. Evol. Comput. 22(4), 623–630 (2017)
Vergnon, R., Dulvy, N.K., Freckleton, R.P.: Niches versus neutrality: uncovering the drivers of diversity in a species-rich community. Ecol. Lett. 12(10), 1079–1090 (2009)
Wessing, S.: Two-stage methods for multimodal optimization. PhD thesis, Universitätsbibliothek Dortmund (2015)
Wright, N.A., Steadman, D.W., Witt, C.C.: Predictable evolution toward flightlessness in volant island birds. Proc. Natl. Acad. Sci. 113(17), 4765–4770 (2016)
Acknowledgements
I would like to thank Alexander Asteroth, Adam Gaier, and Jörg Stork for their feedback. This work received funding from the German Federal Ministry of Education and Research and the Ministry for Culture and Science of the state of North Rhine-Westfalia (research grants 03FH012PX5 and 13FH156IN6).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this chapter
Cite this chapter
Hagg, A. (2021). Phenotypic Niching Using Quality Diversity Algorithms. In: Preuss, M., Epitropakis, M.G., Li, X., Fieldsend, J.E. (eds) Metaheuristics for Finding Multiple Solutions. Natural Computing Series. Springer, Cham. https://doi.org/10.1007/978-3-030-79553-5_12
Download citation
DOI: https://doi.org/10.1007/978-3-030-79553-5_12
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-79552-8
Online ISBN: 978-3-030-79553-5
eBook Packages: Computer ScienceComputer Science (R0)