Thesis (Selection of subject)Thesis (Selection of subject)(version: 368)
Thesis details
   Login via CAS
Evoluční algoritmy pro strukturální učení neuronových sítí
Thesis title in Czech: Evoluční algoritmy pro strukturální učení neuronových sítí
Thesis title in English: Evolutionary algorithms for structural learning of neural networks
Academic year of topic announcement: 2009/2010
Thesis type: diploma thesis
Thesis language: angličtina
Department: Department of Theoretical Computer Science and Mathematical Logic (32-KTIML)
Supervisor: Mgr. Roman Neruda, CSc.
Author: hidden - assigned and confirmed by the Study Dept.
Date of registration: 14.10.2009
Date of assignment: 14.10.2009
Date and time of defence: 13.09.2010 00:00
Date of electronic submission:13.09.2010
Date of proceeded defence: 13.09.2010
Opponents: RNDr. Petra Vidnerová, Ph.D.
 
 
 
Guidelines
Neuroevolution traditionally deals with search space consisting of parameter values of fixed neural topologies only, however it should be able to evolve complete neural structures, as well. Several approaches for structural neuroevolution have one thing in common - a clever and indirect encoding of network topology. Student should investigate approaches by Kitano (grammatical encoding), Gruau (cellular encoding), and Stanley (NEAT), and propose an original algorithm for structural neuroevolution based on his/her research. The approach should be scalable for medium to large size networks and computationally feasible for current computers. A pilot implementation and an experimental evaluation should prove soundness of the approach.
References
Gruau, F., Whitley, D., and Pyeatt, L.: A comparison between cellular encoding and direct encoding for genetic neural networks. In Koza, J. R., Goldberg, D. E., Fogel, D. B., and Riolo, R. L., editors, Genetic Programming 1996: Proceedings of the First Annual Conference, 81?89. Cambridge, MA: MIT Press, 1996.
Kenneth O. Stanley, Risto Miikkulainen: Evolving Neural Networks through Augmenting Topologies. Evolutionary Computation 10(2): 99-127, MIT Press, 2002.
Kenneth O. Stanley, David D'Ambrosio, Jason Gauci: A Hypercube-Based Indirect Encoding for Evolving Large-Scale Neural Networks. Artificial Life journal 15(2), MIT Press, 2009.
Yao, X.: Evolving artificial neural networks. Proceedings of the IEEE, 87(9):1423-1447, 1999.
 
Charles University | Information system of Charles University | http://www.cuni.cz/UKEN-329.html