ABSTRACT
Lexicase parent selection filters the population by considering one random training case at a time, eliminating any individuals with errors for the current case that are worse than the best error in the selection pool, until a single individual remains. This process often stops before considering all training cases, meaning that it will ignore the error values on any cases that were not yet considered. Lexicase selection can therefore select specialist individuals that have poor errors on some training cases, if they have great errors on others and those errors come near the start of the random list of cases used for the parent selection event in question. We hypothesize here that selecting these specialists, which may have poor total error, plays an important role in lexicase selection's observed performance advantages over error-aggregating parent selection methods such as tournament selection, which select specialists much less frequently. We conduct experiments examining this hypothesis, and find that lexicase selection's performance and diversity maintenance degrade when we deprive it of the ability of selecting specialists. These findings help explain the improved performance of lexicase selection compared to tournament selection, and suggest that specialists help drive evolution under lexicase selection toward global solutions.
- Thomas Bäck. 1994. Selective pressure in evolutionary algorithms: a characterization of selection mechanisms. In Evolutionary Computation, 1994. IEEE World Congress on Computational Intelligence., Proceedings of the First IEEE Conference on. 57--62 vol.1.Google ScholarCross Ref
- Tobias Blickle and Lothar Thiele. 1995. A Mathematical Analysis of Tournament Selection. In Proceedings of the 6th International Conference on Genetic Algorithms. Morgan Kaufmann Publishers Inc., San Francisco, CA, USA, 9--16. http://dl.acm.org/citation.cfm?id=645514.658088 Google ScholarDigital Library
- Stefan Forstenlechner, David Fagan, Miguel Nicolau, and Michael O'Neill. 2017. A Grammar Design Pattern for Arbitrary Program Synthesis Problems in Genetic Programming. In EuroGP 2017: Proceedings of the 20th European Conference on Genetic Programming (LNCS), Mauro Castelli, James McDermott, and Lukas Sekanina (Eds.), Vol. 10196. Springer Verlag, Amsterdam, 262--277.Google Scholar
- Stefan Forstenlechner, David Fagan, Miguel Nicolau, and Michael O'Neill. 2018. Extending Program Synthesis Grammars for Grammar-Guided Genetic Programming. In 15th International Conference on Parallel Problem Solving from Nature (LNCS), Anne Auger, Carlos M. Fonseca, Nuno Lourenco, Penousal Machado, Luis Paquete, and Darrell Whitley (Eds.), Vol. 11101. Springer, Coimbra, Portugal, 197--208.Google Scholar
- Stefan Forstenlechner, David Fagan, Miguel Nicolau, and Michael O'Neill. 2018. Towards effective semantic operators for program synthesis in genetic programming. In GECCO '18: Proceedings of the Genetic and Evolutionary Computation Conference. ACM, Kyoto, Japan, 1119--1126. Google ScholarDigital Library
- Stefan Forstenlechner, David Fagan, Miguel Nicolau, and Michael O'Neill. 2018. Towards Understanding and Refining the General Program Synthesis Benchmark Suite with Genetic Programming. In 2018 IEEE Congress on Evolutionary Computation (CEC), Marley Vellasco (Ed.). IEEE, Rio de Janeiro, Brazil.Google Scholar
- Thomas Helmuth, Nicholas Freitag McPhee, Edward Pantridge, and Lee Spector. 2017. Improving Generalization of Evolved Programs Through Automatic Simplification. In Proceedings of the Genetic and Evolutionary Computation Conference (GECCO '17). ACM, Berlin, Germany, 937--944. Google ScholarDigital Library
- Thomas Helmuth, Nicholas Freitag McPhee, and Lee Spector. 2015. Lexicase Selection For Program Synthesis: A Diversity Analysis. In Genetic Programming Theory and Practice XIII (Genetic and Evolutionary Computation), Rick Riolo, William P. Worzel, M. Kotanchek, and A. Kordon (Eds.). Springer, Ann Arbor, USA.Google Scholar
- Thomas Helmuth, Nicholas Freitag McPhee, and Lee Spector. 2016. Effects of Lexicase and Tournament Selection on Diversity Recovery and Maintenance. In GECCO '16 Companion: Proceedings of the Companion Publication of the 2016 Annual Conference on Genetic and Evolutionary Computation. ACM, Denver, Colorado, USA, 983--990. Google ScholarDigital Library
- Thomas Helmuth, Nicholas Freitag McPhee, and Lee Spector. 2016. The Impact of Hyperselection on Lexicase Selection. In GECCO '16: Proceedings of the 2016 Annual Conference on Genetic and Evolutionary Computation, Tobias Friedrich (Ed.). ACM, Denver, USA, 717--724. Google ScholarDigital Library
- Thomas Helmuth and Lee Spector. 2013. Evolving a digital multiplier with the PushGP genetic programming system. In GECCO '13 Companion: Proceeding of the fifteenth annual conference companion on Genetic and evolutionary computation conference companion. ACM, Amsterdam, The Netherlands, 1627--1634. Google ScholarDigital Library
- Thomas Helmuth and Lee Spector. 2015. General Program Synthesis Benchmark Suite. In GECCO '15: Proceedings of the 2015 Annual Conference on Genetic and Evolutionary Computation. ACM, Madrid, Spain, 1039--1046. Google ScholarDigital Library
- Thomas Helmuth, Lee Spector, and James Matheson. 2015. Solving Uncompromising Problems with Lexicase Selection. IEEE Transactions on Evolutionary Computation 19, 5 (Oct. 2015), 630--643.Google ScholarDigital Library
- Thomas M. Helmuth. 2015. General Program Synthesis from Examples Using Genetic Programming with Parent Selection Based on Random Lexicographic Orderings of Test Cases. Ph.D. Dissertation. College of Information and Computer Sciences, University of Massachusetts Amherst, USA. https://web.cs.umass.edu/publication/docs/2015/UM-CS-PhD-2015-005.pdfGoogle Scholar
- David Jackson. 2010. Promoting Phenotypic Diversity in Genetic Programming. In PPSN 2010 11th International Conference on Parallel Problem Solving From Nature (Lecture Notes in Computer Science), Robert Schaefer, Carlos Cotta, Joanna Kolodziej, and Guenter Rudolph (Eds.), Vol. 6239. Springer, Krakow, Poland, 472--481. Google ScholarDigital Library
- William La Cava, Thomas Helmuth, Lee Spector, and Jason H. Moore. 2018. A probabilistic and multi-objective analysis of lexicase selection and epsilon-lexicase selection. Evolutionary Computation (2018). Forthcoming.Google Scholar
- William La Cava, Lee Spector, and Kourosh Danai. 2016. Epsilon-lexicase Selection for Regression. In GECCO '16: Proceedings of the 2016 Annual Conference on Genetic and Evolutionary Computation, Tobias Friedrich (Ed.). ACM, Denver, USA, 741--748. Google ScholarDigital Library
- Pawel Liskowski, Krzysztof Krawiec, Thomas Helmuth, and Lee Spector. 2015. Comparison of Semantic-aware Selection Methods in Genetic Programming. In GECCO 2015 Semantic Methods in Genetic Programming (SMGP'15) Workshop. ACM, Madrid, Spain, 1301--1307. Google ScholarDigital Library
- Nicholas Freitag McPhee, David Donatucci, and Thomas Helmuth. 2015. Using Graph Databases to Explore Genetic Programming Run Dynamics. In Genetic Programming Theory and Practice XIII (Genetic and Evolutionary Computation). Springer, Ann Arbor, USA. http://www.springer.com/us/book/9783319342214Google Scholar
- Nicholas Freitag McPhee, Thomas Helmuth, and Lee Spector. 2017. Using Algorithm Configuration Tools to Optimize Genetic Programming Parameters: A Case Study. In Proceedings of the Genetic and Evolutionary Computation Conference Companion (GECCO '17). ACM, Berlin, Germany, 243--244. Google ScholarDigital Library
- Blossom Metevier, Anil Kumar Saini, and Lee Spector. 2019. Lexicase Selection Beyond Genetic Programming. Springer International Publishing, Cham, 123--136.Google Scholar
- Jared M. Moore and Adam Stanton. 2017. Lexicase selection outperforms previous strategies for incremental evolution of virtual creature controllers. Proceedings of the European Conference on Artificial Life (2017), 290--297.Google Scholar
- Edward Pantridge, Thomas Helmuth, Nicholas Freitag McPhee, and Lee Spector. 2018. Specialization and Elitism in Lexicase and Tournament Selection. In Proceedings of the Genetic and Evolutionary Computation Conference Companion (GECCO '18). ACM, New York, NY, USA, 1914--1917. Google ScholarDigital Library
- Edward Pantridge and Lee Spector. 2017. PyshGP: PushGP in Python. In Proceedings of the Genetic and Evolutionary Computation Conference Companion (GECCO '17). ACM, Berlin, Germany, 1255--1262. Google ScholarDigital Library
- Christopher D. Rosin. 2018. Stepping Stones to Inductive Synthesis of Low-Level Looping Programs. CoRR abs/1811.10665 (2018). arXiv:1811.10665 http://arxiv.org/abs/1811.10665Google Scholar
- Lee Spector. 2012. Assessment of Problem Modality by Differential Performance of Lexicase Selection in Genetic Programming: A Preliminary Report. In 1st workshop on Understanding Problems (GECCO-UP), Kent McClymont and Ed Keedwell (Eds.). ACM, Philadelphia, Pennsylvania, USA, 401--408. Google ScholarDigital Library
- Lee Spector, Jon Klein, and Maarten Keijzer. 2005. The Push3 execution stack and the evolution of control. In GECCO 2005: Proceedings of the 2005 conference on Genetic and evolutionary computation, Vol. 2. ACM Press, Washington DC, USA, 1689--1696. Google ScholarDigital Library
- Lee Spector, William La Cava, Saul Shanabrook, Thomas Helmuth, and Edward Pantridge. 2018. Relaxations of Lexicase Parent Selection. In Genetic Programming Theory and Practice XV, Wolfgang Banzhaf, Randal S. Olson, William Tozier, and Rick Riolo (Eds.). Springer International Publishing, Cham, 105--120.Google Scholar
- Lee Spector and Alan Robinson. 2002. Genetic Programming and Autoconstructive Evolution with the Push Programming Language. Genetic Programming and Evolvable Machines 3, 1 (March 2002), 7--40. Google ScholarDigital Library
Index Terms
- Lexicase selection of specialists
Recommendations
Lexicase selection in learning classifier systems
GECCO '19: Proceedings of the Genetic and Evolutionary Computation ConferenceThe lexicase parent selection method selects parents by considering performance on individual data points in random order instead of using a fitness function based on an aggregated data accuracy. While the method has demonstrated promise in genetic ...
The Impact of Hyperselection on Lexicase Selection
GECCO '16: Proceedings of the Genetic and Evolutionary Computation Conference 2016Lexicase selection is a parent selection method that has been shown to improve the problem solving power of genetic programming over a range of problems. Previous work has shown that it can also produce hyperselection events, in which a single ...
On the importance of specialists for lexicase selection
AbstractLexicase parent selection filters the population by considering one random training case at a time, eliminating any individual with an error for the current case that is worse than the best error of any individual in the selection pool, until a ...
Comments