Abstract
The paper explores the issue of learning multiple competing tasks in the domain of artificial evolution. In particular, a robot is trained so as to be able to perform two different tasks at the same time, namely a gradient following and rough terrain avoidance behaviours. It is shown that, if the controller is trained to learn two tasks of different difficulty, then the robot performance is higher if the most difficult task is learnt first, before the combined learning of both tasks. An explanation to this superiority is also discussed, in comparison with previous results.
Keywords
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
Let us observe that such parameters have been set to the present values after a careful analysis of some generic simulations and by a test and error procedure.
- 2.
More precisely, one network composed by 9 inputs, 5 hidden neurons and 4 outputs, has only the inputs able to measure the local heights and the second one, formed by 8 inputs, 5 hidden neurons and 4 outputs, to detect the dangerous zones.
References
Beaumont, M.A.: Evolution of optimal behaviour in networks of boolean automata. J. Theor. Biol. 165, 455–476 (1993)
Blondel, V.D., Guillaume, J.-L., Lambiotte, R., Lefebvre, E.: Fast unfolding of communities in large networks. J. Stat. Mech: Theory Exp. 2008(10), P10008 (2008)
Calabretta, R., Di Ferdinando, A., Parisi, D., Keil, F.C.: How to learn multiple tasks. Biol. Theor. 3(1), 30–41 (2008)
Deb, K.: Multi-Objective Optimization using Evolutionary Algorithms. John Wiley & Sons Ltd, Chichester (2008)
Dorigo, M., Colombetti, M.: Robot Shaping: An Experiment in Behavior Engineering. The MIT Press, Cambridge (1998)
Goldberg, D.E.: Genetic Algorithms in Search, Optimization, and Machine Learning. Addison-Wesley, Boston (1989)
Hollander, M., Wolfe, D.A., Chicken, E.: Nonparametric statistical methods, vol. 751. John Wiley & Sons, New York (2013)
Hoos, H.H., Stützle, T.: Stochastic local search: Foundations & applications. Elsevier (2004)
Kashtan, N., Alon, U.: Spontaneous evolution of modularity and network motifs. Proc. Nat. Acad. Sci. U.S.A. 102(39), 13773–13778 (2005)
MacKay, D.: Information Theory, Inference, and Learning Algorithms. Cambridge University Press, Cambridge (2005)
Nicolay, D., Carletti, T.: Neural networks learning: some preliminary results on heuristic methods and applications. In: Perotti, A., Di Caro, L. (eds.) DWAI@AI*IA, volume 1126 of CEUR Workshop Proceedings, pp. 30–40 (2013). CEUR-WS.org
Nolfi, S., Floreano, D.: Evolutionary robotics. The MIT Press, Cambridge (2000)
Peretto, P.: An introduction to the modeling of neural networks, vol. Alea Saclay. Cambridge University Press, Cambridge (1992)
Rojas, R.: Neural networks: A systematic introduction. Springer, Berlin (1996)
Villani, M., Filisetti, A., Benedettini, S., Roli, A., Lane, D.A., Serra, R.: The detection of intermediate-level emergent structures and patterns. In: Advances in Artificial Life, ECAL, vol. 12, pp. 372–378 (2013)
Acknowledgements
This research used computational resources of the “Plateforme Technologique de Calcul Intensif (PTCI)” located at the University of Namur, Belgium, which is supported by the F.R.S.-FNRS.
This paper presents research results of the Belgian Network DYSCO (Dynamical Systems, Control, and Optimisation), funded by the Interuniversity Attraction Poles Programme, initiated by the Belgian State, Science Policy Office. The scientific responsibility rests with its authors.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer International Publishing Switzerland
About this paper
Cite this paper
Nicolay, D., Roli, A., Carletti, T. (2014). Learning Multiple Conflicting Tasks with Artificial Evolution. In: Pizzuti, C., Spezzano, G. (eds) Advances in Artificial Life and Evolutionary Computation. WIVACE 2014. Communications in Computer and Information Science, vol 445. Springer, Cham. https://doi.org/10.1007/978-3-319-12745-3_11
Download citation
DOI: https://doi.org/10.1007/978-3-319-12745-3_11
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-12744-6
Online ISBN: 978-3-319-12745-3
eBook Packages: Computer ScienceComputer Science (R0)