Abstract
We present a study looking at the effect of a priori domain knowledge on an EA fitness function. Our experiment has two aims: (1) applying an existing NSGA-II framework for GP with PDL to the cartpole problem—applying GP & PDL to cartpole and a purely behavioral problem for the first time—and (2) contrasting two multi-objective fitness functions: one with high and the other with low a priori domain knowledge. In our experiment we created two populations with an EA, varying in the number of objectives use for the fitness function, 2 objective criteria to represent low a priori knowledge and 3 to represent high. With fitness functions tailored to find specifically prescribed solutions we expect greater discriminating power and more feedback to an evolutionary process. This comes at the cost of excluding some unexpected solutions from the evolutionary process and placing a greater burden on the designer. We address the question: how large is the disadvantage for the low a priori fitness function in a worst-case scenario, where innovative solutions will not enhance performance. This question is interesting because we would prefer to guide EA with simple, easy to create and understand, objective criteria rather than complex and highly specific criteria. Understanding any associated penalty for using simple, easy to create fitness functions, is crucial in assessing how much effort and should be put into designing objective criteria.
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.
A video of the representatives from both groups is available at https://youtu.be/99S11Kr9vRs.
- 2.
- 3.
References
Steels, L.: Mathematical analysis of behavior systems. In: Proceedings of PerAc ‘94. From Perception to Action, pp. 88–95. IEEE Computer Society Press (1994)
Shannon, P., Nehaniv, C L.: Evolving robot controllers in PDL using genetic programming. In: 2011 IEEE Symposium on Artificial Life (IEEE ALIFE), pp. 92–99. IEEE (2011)
Shannon, P.D., Nehaniv, C.L., Phon-Amnuaisuk, S.: Enhancing exploration and exploitation of NSGA-II with GP and PDL. In: Tan, Y., Takagi, H., Shi, Y. (eds.) ICSI 2017. LNCS, vol. 10385, pp. 349–361. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-61824-1_38
Koza, J.R.: Genetic Programming: on the Programming of Computers by Means of Natural Selection. MIT Press, Cambridge (1992)
Nelson, A.L., et al.: Fitness functions in evolutionary robotics: a survey and analysis. Robot. Auton. Syst. 57(4), 345–370 (2009)
Brooks, R.: A robust layered control system for a mobile robot. IEEE J. Robot. Autom. 2(1), 14–23 (1986)
Sutton, R.S., Barto, A.G.: Reinforcement Learning: An Introduction. MIT Press, Cambridge (2011)
Schlottmann, E., et al.: A modular design approach towards behaviour oriented robotics (1997)
Birk, A., et al.: Programming with behavior-processes. Robot. Auton. Syst. 39(3–4), 115–127 (2002)
Barto, A.G., et al.: Neuronlike adaptive elements that can solve difficult learning control problems. In: IEEE Transactions on Systems, Man, and Cybernetics 5, pp. 834–846 (1983)
Koza, J.R., Keane, M.A.: Genetic breeding of non-linear optimal control strategies for broom balancing. In: Bensoussan, A., Lions, J.L. (eds.) Analysis and Optimization of Systes, pp. 47–56. Springer, Berlin (1990). https://doi.org/10.1007/BFb0120027
Fortin, F.-A., et al.: DEAP: evolutionary algorithms made easy. J. Mach. Learn. Res. 13(1), 2171–2175 (2012)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2021 Springer Nature Switzerland AG
About this paper
Cite this paper
Shannon, P.D., Nehaniv, C.L., Phon-Amnuaisuk, S. (2021). Cartpole Problem with PDL and GP Using Multi-objective Fitness Functions Differing in a Priori Knowledge. In: Chomphuwiset, P., Kim, J., Pawara, P. (eds) Multi-disciplinary Trends in Artificial Intelligence. MIWAI 2021. Lecture Notes in Computer Science(), vol 12832. Springer, Cham. https://doi.org/10.1007/978-3-030-80253-0_10
Download citation
DOI: https://doi.org/10.1007/978-3-030-80253-0_10
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-80252-3
Online ISBN: 978-3-030-80253-0
eBook Packages: Computer ScienceComputer Science (R0)