ABSTRACT
During the initialization step, a genetic programming (GP) system traditionally creates a population of completely random programs to populate the initial population. These programs almost always perform poorly in terms of their total error---some might not even output the correct data type. In this paper, we present new methods for initialization that attempt to generate programs that are somewhat relevant to the problem being solved and/or increase the initial diversity (both error and behavioral diversity) of the population prior to the GP run. By seeding the population---and thereby eliminating worthless programs and increasing the initial diversity of the population---we hope to improve the performance of the GP system. Here, we present two novel techniques for initialization (Lexicase Seeding and Pareto Seeding) and compare them to a previous method (Enforced Diverse Populations) and traditional, non-seeded initialization. Surprisingly, we found that none of the initialization methods result in significant differences in problem-solving performance or population diversity across five program synthesis benchmark problems. We attribute this lack of difference to our use of lexicase selection, which seems to rapidly converge on similar populations regardless of initialization method.
- Stefan Forstenlechner, David Fagan, Miguel Nicolau, and Michael O'Neill. 2017. A Grammar Design Pattern for Arbitrary Program Synthesis Problems in Genetic Programming. In 20th European Conference on Genetic Programming. In press.Google ScholarCross Ref
- 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). Springer, Ann Arbor, USA. DOI: http://dx.doi.org/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 2016 on Genetic and Evolutionary Computation Conference Companion. ACM, Denver, Colorado, USA, 983--990. DOI:http://dx.doi.org/ 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 on Genetic and Evolutionary Computation Conference, Tobias Friedrich (Ed.). ACM. Denver, USA, 717--724. DOI:http://dx.doi.org/ Google ScholarDigital Library
- Thomas Helmuth and Lee Spector. 2015. General Program Synthesis Benchmark Suite. In GECCO '15: Proceedings of the 2015 on Genetic and Evolutionary Computation Conference. ACM, Madrid, Spain, 1039--1046. DOI:http://dx.doi.org/ 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. DOI:http://dx.doi.org/Google ScholarDigital Library
- Thomas Helmuth, Lee Spector, Nicholas Freitag McPhee, and Saul Shanabrook. 2016. Linear Genomes for Structured Programs. In Genetic Programming Theory and Practice XIV (Genetic and Evolutionary Computation). Springer, Ann Arbor. USA.Google Scholar
- David Jackson. 2010. Phenotypic Diversity in Initial Genetic Programming Populations. In Proceedings of the 13th European Conference on Genetic Programming, EuroGP 2010 (LNCS), Vol. 6021. Springer, Istanbul, 98--109. DOI:http://dx.doi.org/ Google ScholarDigital Library
- John R. Koza. 1992. Genetic Programming: On the Programming of Computers by Means of Natural Selection. MIT Press, Cambridge, MA, USA. http://mitpress.mit.edu/books/genetic-programming Google ScholarDigital Library
- Sean Luke. 2000. Two Fast Tree-Creation Algorithms for Genetic Programming. IEEE Transactions on Evolutionary Computation 4, 3 (Sept. 2000), 274--283. http://ieeexplore.ieee.org/iel5/4235/18897/00873237.pdf Google ScholarDigital Library
- Riccardo Poli, William B. Langdon, and Nicholas Freitag McPhee. 2008. A field guide to genetic programming. Published via http://lulu.com and freely available at http://www.gp-field-guide.org.uk. http://www.gp-field-guide.org.uk (With contributions by J. R. Koza). Google ScholarDigital Library
- Michael Schmidt and Hod Lipson. 2010. Age-Fitness Pareto Optimization. In Genetic Programming Theory and Practice VIII, Rick Riolo, Trent McConaghy. and Ekaterina Vladislavleva (Eds.). Genetic and Evolutionary Computation. Vol. 8. Springer, Ann Arbor, USA, Chapter 8, 129--146. http://www.springer.com/computer/ai/book/978-1-4419-7746-5 Google ScholarDigital Library
- Lee Spector. 2001. Autoconstructive Evolution: Push, PushGP, and Push-pop. In Proceedings of the Genetic and Evolutionary Computation Conference (GECCO-2001). Morgan Kaufmann, San Francisco, California, USA, 137--146. http://hampshire.edu/lspector/pubs/ace.pdfGoogle Scholar
- 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. ACM Press, Washington DC, USA. 1689--1696. DOI:http://dx.doi.org/ Google ScholarDigital Library
Index Terms
- A comparison of semantic-based initialization methods for genetic programming
Recommendations
An Efficient Structural Diversity Technique for Genetic Programming
GECCO '15: Proceedings of the 2015 Annual Conference on Genetic and Evolutionary ComputationGenetic diversity plays an important role in avoiding premature convergence, which is a phenomenon that stifles the search effectiveness of evolutionary algorithms. However, approaches that avoid premature convergence by maintaining genetic diversity ...
Grammatically uniform population initialization for grammar-guided genetic programming
AbstractThe initial population distribution is an essential issue in evolutionary computation performance. Population initialization methods for grammar-guided genetic programming have some difficulties generating a representative sample of the search ...
Diverse partner selection with brood recombination in genetic programming
Graphical abstractDisplay Omitted
Highlights- Optimal parent selection during crossover in genetic programming.
- Exploitation ...
AbstractThe ultimate goal of learning algorithms is to find the best solution from a search space without testing each and every solution available in the search space. During the evolution process new solutions (children) are produced from ...
Comments