ITU

Enhancing Divergent Search through Extinction Events

Research output: Conference Article in Proceeding or Book/Report chapter › Article in proceedings › Research › peer-review

Standard

Enhancing Divergent Search through Extinction Events. / Lehman, Joel; Miikkulainen, Risto.

Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2015): GECCO '15. New York, NY, USA : Association for Computing Machinery, 2015. p. 951-958.

Research output: Conference Article in Proceeding or Book/Report chapter › Article in proceedings › Research › peer-review

Harvard

Lehman, J & Miikkulainen, R 2015, Enhancing Divergent Search through Extinction Events. in Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2015): GECCO '15. Association for Computing Machinery, New York, NY, USA, pp. 951-958. https://doi.org/10.1145/2739480.2754668

APA

Lehman, J., & Miikkulainen, R. (2015). Enhancing Divergent Search through Extinction Events. In Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2015): GECCO '15 (pp. 951-958). Association for Computing Machinery. https://doi.org/10.1145/2739480.2754668

Vancouver

Lehman J, Miikkulainen R. Enhancing Divergent Search through Extinction Events. In Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2015): GECCO '15. New York, NY, USA: Association for Computing Machinery. 2015. p. 951-958 https://doi.org/10.1145/2739480.2754668

Author

Lehman, Joel ; Miikkulainen, Risto. / Enhancing Divergent Search through Extinction Events. Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2015): GECCO '15. New York, NY, USA : Association for Computing Machinery, 2015. pp. 951-958

Bibtex

@inproceedings{f4dfbf4ef2ce496a94723193a3abdd66,
title = "Enhancing Divergent Search through Extinction Events",
abstract = "A challenge in evolutionary computation is to create representations as evolvable as those in natural evolution. This paper hypothesizes that extinction events, i.e. mass extinctions, can significantly increase evolvability, but only when combined with a divergent search algorithm, i.e. a search driven towards diversity (instead of optimality). Extinctions amplify diversity-generation by creating unpredictable evolutionary bottlenecks. Persisting through multiple such bottlenecks is more likely for lineages that diversify across many niches, resulting in indirect selection pressure for the capacity to evolve. This hypothesis is tested through experiments in two evolutionary robotics domains. The results show that combining extinction events with divergent search increases evolvability, while combining them with convergent search offers no similar benefit. The conclusion is that extinction events may provide a simple and effective mechanism to enhance performance of divergent search algorithms.",
author = "Joel Lehman and Risto Miikkulainen",
year = "2015",
doi = "10.1145/2739480.2754668",
language = "English",
isbn = "978-1-4503-3472-3",
pages = "951--958",
booktitle = "Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2015)",
publisher = "Association for Computing Machinery",
address = "United States",

}

RIS

TY - GEN

T1 - Enhancing Divergent Search through Extinction Events

AU - Lehman, Joel

AU - Miikkulainen, Risto

PY - 2015

Y1 - 2015

N2 - A challenge in evolutionary computation is to create representations as evolvable as those in natural evolution. This paper hypothesizes that extinction events, i.e. mass extinctions, can significantly increase evolvability, but only when combined with a divergent search algorithm, i.e. a search driven towards diversity (instead of optimality). Extinctions amplify diversity-generation by creating unpredictable evolutionary bottlenecks. Persisting through multiple such bottlenecks is more likely for lineages that diversify across many niches, resulting in indirect selection pressure for the capacity to evolve. This hypothesis is tested through experiments in two evolutionary robotics domains. The results show that combining extinction events with divergent search increases evolvability, while combining them with convergent search offers no similar benefit. The conclusion is that extinction events may provide a simple and effective mechanism to enhance performance of divergent search algorithms.

AB - A challenge in evolutionary computation is to create representations as evolvable as those in natural evolution. This paper hypothesizes that extinction events, i.e. mass extinctions, can significantly increase evolvability, but only when combined with a divergent search algorithm, i.e. a search driven towards diversity (instead of optimality). Extinctions amplify diversity-generation by creating unpredictable evolutionary bottlenecks. Persisting through multiple such bottlenecks is more likely for lineages that diversify across many niches, resulting in indirect selection pressure for the capacity to evolve. This hypothesis is tested through experiments in two evolutionary robotics domains. The results show that combining extinction events with divergent search increases evolvability, while combining them with convergent search offers no similar benefit. The conclusion is that extinction events may provide a simple and effective mechanism to enhance performance of divergent search algorithms.

U2 - 10.1145/2739480.2754668

DO - 10.1145/2739480.2754668

M3 - Article in proceedings

SN - 978-1-4503-3472-3

SP - 951

EP - 958

BT - Proceedings of the Genetic and Evolutionary Computation Conference (GECCO 2015)

PB - Association for Computing Machinery

CY - New York, NY, USA

ER -

ID: 80652049