ABSTRACT
We propose a new form of Cartesian Genetic Programming (CGP) that develops into a computational network capable of learning. The developed network architecture is inspired by the brain. When the genetically encoded programs are run, a networks develops consisting of neurons, dendrites, axons, and synapses which can grow, change or die. We have tested this approach on the task of learning how to play checkers. The novelty of the research lies mainly in two aspects: Firstly, chromosomes are evolved that encode programs rather than the network directly and when these programs are executed they build networks which appear to be capable of learning and improving their performance over time solely through interaction with the environment. Secondly, we show that we can obtain learning programs much quicker through co-evolution in comparison to the evolution of agents against a minimax based checkers program. Also, co-evolved agents show significantly increased learning capabilities compared to those that were evolved to play against a minimax-based opponent.
- A. Cangelosi, S. Nolfi, and D. Parisi. Cell division and migration in a 'genotype' for neural networks. Network-Computation in Neural Systems, 5:497--515, 1994.Google ScholarCross Ref
- F. Dalaert and R. Beer. Towards an evolvable model of development for autonomous agent synthesis. In Brooks, R. and Maes, P. eds. Proceedings of the Fourth Conference on Artificial Life. MIT Press, 1994.Google Scholar
- R. Dawkins and J. R. Krebs. Arms races between and within species. In Proceedings of the Royal Society of London Series B, volume 205, page 489U511, 1979.Google Scholar
- D. Federici. Evolving developing spiking neural networks. In Proceedings of CEC 2005 IEEE Congress on Evolutionary Computation, pages 543--550, 2005.Google ScholarCross Ref
- D. Fogel. Blondie24: Playing at the Edge of AI. Academic Press, London, UK, 2002. Google ScholarDigital Library
- F. Gruau. Automatic definition of modular neural networks. Adaptive Behaviour, 3:151--183, 1994. Google ScholarDigital Library
- W. Hillis. Co-evolving parasites improve simulated evolution as an optimization procedure. Artificial life 2, pages 313--324, 1991.Google Scholar
- N. Jacobi. Harnessing Morphogenesis, Cognitive Science Research Paper 423, COGS. University of Sussex, 1995.Google Scholar
- G. Kendall and G. Whitwell. An evolutionary approach for the tuning of a chess evaluation function using population dynamics. In IEEE. CEC. 2001, pages 995--1002, 2001.Google Scholar
- G. Khan, J. Miller, and D. Halliday. Coevolution of intelligent agents using cartesian genetic programming. In Proc. GECCO, pages 269--276, 2007. Google ScholarDigital Library
- A. Lubberts and R. Miikkulainen. Co-evolving a go-playing neural network. in Coevolution: Turning Adaptive Algorithms upon Themselves, Belew R. and Juille H (eds.), pages 14--19, 2001.Google Scholar
- J. Miller, D. Job, and V. Vassilev. Principles in the evolutionary design of digital circuits -- part i. Journal of Genetic Programming and Evolvable Machines, 1(2):259--288, 2000. Google ScholarDigital Library
- J. F. Miller and P. Thomson. Cartesian genetic programming. In Proc. EuroGP, volume 1802 of LNCS, pages 121--132, 2000. Google ScholarDigital Library
- D. Moriarty and R. Miikulainen. Discovering complex othello strategies through evolutionary neural networks. Connection Science, 7(3-4):195--209, 1995.Google ScholarCross Ref
- S. Nolfi and D. Floreano. Co-evolving predator and prey robots: Do 'arm races' arise in artificial evolution? Artificial Life, 4:311--335, 1998. Google ScholarDigital Library
- S. Nolfi, O. Miglino, and D. Parisi. Phenotypic plasticity in evolving neural networks. in gaussier, d.p, and nicoud, j.d., eds. In Proceedings of the International Conference from perception to action. IEEE Press, 1994.Google ScholarCross Ref
- J. Paredis. Coevolutionary constraint satisfaction. In Proceedings of the third international conference on parallel problem solving from nature, Springer-Verlag, volume 866, pages 46--55, 1994. Google ScholarDigital Library
- J. Paredis. Coevolutionary computation. Artificial Life, 2(4):355--375, 1995. Google ScholarDigital Library
- J. Pollack, A. Blair, and M. Land. Coevolution of a backgammon player. In In: Langton, C.(ed),Proceedings artificial life 5. MIT Press.Google Scholar
- D. Roggen, D. Federici, and D. Floreano. Evolutionary morphogenesis for multi-cellular systems. Journal of Genetic Programming and Evolvable Machines, 8:61--96, 2007. Google ScholarDigital Library
- C. D. Rosin. Coevolutionary search among adversaries. Ph.D. thesis, University of California, San Diego., 1997. Google ScholarDigital Library
- A. Rust, R. Adams, and B. H. Evolutionary neural topiary: Growing and sculpting artificial neurons to order. In Proc. of the 7th Int. Conf. on the Simulation and synthesis of Living Systems (ALife VII), pages 146--150. MIT Press, 2000.Google Scholar
- A. G. Rust, R. Adams, S. George, and H. Bolouri. Activity-based pruning in developmental artificial neural networks. In Proc. of the European Conf. on Artificial Life (ECAL'97), pages 224--233. MIT Press, 1997.Google Scholar
- J. Schaeffer. One Jump Ahead: Challenging Human Supremacy in Checkers. Springer, Berlin, 1996. Google ScholarDigital Library
- G. Shepherd. The synaptic organization of the brain. Oxford Press, 1990.Google Scholar
- A. Van Ooyen and J. Pelt. Activity-dependent outgrowth of neurons and overshoot phenomena in developing neural networks. Journal of Theoretical Biology, 167:27--43, 1994.Google ScholarCross Ref
Index Terms
- Evolution of cartesian genetic programs capable of learning
Recommendations
Developing neural structure of two agents that play checkers using cartesian genetic programming
GECCO '08: Proceedings of the 10th annual conference companion on Genetic and evolutionary computationA developmental model of neural network is presented and evaluated in the game of Checkers. The network is developed using cartesian genetic programs (CGP) as genotypes. Two agents are provided with this network and allowed to co-evolve untill they ...
Coevolution of intelligent agents using cartesian genetic programming
GECCO '07: Proceedings of the 9th annual conference on Genetic and evolutionary computationA coevolutionary competitive learning environment for two antagonistic agents is presented. The agents are controlled by a new kind of computational network based on a compartmentalised model of neurons. We have taken the view that the genetic basis of ...
Evolution of cartesian genetic programs for development of learning neural architecture
Although artificial neural networks have taken their inspiration from natural neurological systems, they have largely ignored the genetic basis of neural functions. Indeed, evolutionary approaches have mainly assumed that neural learning is associated ...
Comments