ABSTRACT
Solving complex, real-world problems with genetic programming (GP) can require extensive computing resources. However, the highly parallel nature of GP facilitates using a large number of resources simultaneously, which can significantly reduce the elapsed wall clock time per GP run. This paper explores the performance characteristics of an MPI version of the Genetic Programming Environment for FIFTH (GPE5) on a high performance computing cluster. The implementation is based on the island model with each node running the GP algorithm asynchronously. In particular, we examine the effect of several configurable properties of the system including the ratio of migration to crossover, the migration cycle of programs between nodes, and the number of processors used. The problems employed in the study were selected from the fields of symbolic regression, finite algebra, and digital signal processing.
- Andre, D. and Koza, J.R. A parallel implementation of genetic programming that achieves super--linear performance. Arabnia, H.R. ed. Proceedings of the International Conference on Parallel and Distributed Processing Techniques and Applications, CSREA, Sunnyvale, 1996, 1163--1174.Google Scholar
- Daida, J.M., Bertram, R.R., A,J. and Stanhope, S.A. Analysis of Single-Node (Building) Blocks in Genetic Programming. Spector, L., Langdon, n.W.B., O'Reilly, n.U.-M. and Angeline, n.P.J. eds. Advances in Genetic Programming 3, MIT Press, Cambridge, MA, USA, 1999, 217--241. Google ScholarDigital Library
- Daida, J.M., Bertram, R.R., Stanhope, S.A., Khoo, J.C., Chaudhary, S.A., Chaudhri, O.A. and A, J. What Makes a Problem GP-Hard? Analysis of a Tunably Difficult Problem in Genetic Programming Genetic Programming and Evolvable Machines, 2001, 165--191. Google ScholarDigital Library
- Fernandez, F., Tomassini, M. and Vanneschi, L. An Empirical Study of Multipopulation Genetic Programming Genetic Programming and Evolvable Machines, 2003, 21--51. Google ScholarDigital Library
- Fernandez, F., Tomassini, M., Vanneschi, L. and Bucher, L. A Distributed Computing Environment for Genetic Programming using MPI. Dongarra, J.J., Kacsuk, n.P. and Podhorszki, n.N. eds. Recent advances in parallel virtual machine and message passing interface: 7th European PVM\slash MPI Users' Group Meeting, Springer-Verlag, Balatonfured, Hungary, 2000, 322--329. Google ScholarDigital Library
- Folino, G., Pizzuti, C. and Spezzano, G. A Scalable Cellular Implementation of Parallel Genetic Programming IEEE Transactions on Evolutionary Computation, 2003, 37--53. Google ScholarDigital Library
- Holladay, K., Robbins, K. and Ronne, J.v. FIFTH": A stack based GP language for vector processing. Ebner, O.N., Ekart, Vanneschi and Esparcia Alcazar ed. EuroGP 2007, Springer, Valencia, Spain, 2007. Google ScholarDigital Library
- Holladay, K.L. and Robbins, K.A., Evolution of Signal Processing Algorithms using Vector Based Genetic Programming. in Digital Signal Processing, 2007 15th International Conference on, (Cardiff, Wales, 2007), 503--506.Google ScholarCross Ref
- Koza, J.R. Genetic Programming II: Automatic Discovery of Reusable Programs. MIT Press, Cambridge, 1994. Google ScholarDigital Library
- Koza, J.R. Genetic Programming: On the Programming of Computers by Means of Natural Selection. MIT Press, Cambridge, 1992. Google ScholarDigital Library
- Potter, M.A. and Jong, K.A.D. Cooperative Coevolution: An Architecture for Evolving Coadapted Subcomponents. Evolutionary Computation, 8 (1). 1--29. Google ScholarDigital Library
- Snir, M. MPI--The Complete Reference. MIT Press, Cambridge, Mass., 1998. Google ScholarDigital Library
- Spector, L., Clark, D.M., Lindsay, I., Barr, B. and Klein, J. Genetic programming for finite algebras Proceedings of the 10th annual conference on Genetic and evolutionary computation, ACM, Atlanta, GA, USA, 2008. Google ScholarDigital Library
- Spector, L., Klein, J. and Keijzer, M., The Push3 Execution Stack and the Evolution of Control. in GECCO 2005: Proceedings of the 2005 conference on Genetic and evolutionary computation, (Washington DC, USA, 2005), ACM SIGEVO (formerly ISGEC) ACM Press New York, NY, 10286-1405, USA, 1689--1696. Google ScholarDigital Library
- Tchernev, E. Stack-Correct Crossover Methods in Genetic Programming. Cant\'u-Paz, E. ed. Late Breaking Papers at the Genetic and Evolutionary Computation Conference (GECCO-2002), AAAI 445 Burgess Drive, Menlo Park, CA 94025, New York, NY, 2002, 443--449.Google Scholar
Recommendations
Neural network crossover in genetic algorithms using genetic programming
AbstractThe use of genetic algorithms (GAs) to evolve neural network (NN) weights has risen in popularity in recent years, particularly when used together with gradient descent as a mutation operator. However, crossover operators are often omitted from ...
Exact Schema Theory and Markov Chain Models for Genetic Programming and Variable-length Genetic Algorithms with Homologous Crossover
Genetic Programming (GP) homologous crossovers are a group of operators, including GP one-point crossover and GP uniform crossover, where the offspring are created preserving the position of the genetic material taken from the parents. In this paper we ...
Performance improvement in genetic programming using modified crossover and node mutation
GECCO '13 Companion: Proceedings of the 15th annual conference companion on Genetic and evolutionary computationDuring the evolution of solutions using Genetic Programming (GP) there is generally an increase in average tree size without a corresponding increase in fitness'a phenomenon commonly referred to as bloat. Bloating increases time to find the best ...
Comments