On the search for new learning rules for ANNs
- 33 Downloads
In this paper, we present a framework where a learning rule can be optimized within a parametric learning rule space. We define what we callparametric learning rules and present a theoretical study of theirgeneralization properties when estimated from a set of learning tasks and tested over another set of tasks. We corroborate the results of this study with practical experiments.
Unable to display preview. Download preview PDF.
- D. Chalmers. The evolution of learning: An experiment in genetic connectionism, inProceedings of the 1990 Connectionist Models Summer School, J. Hinton, ed., San Mateo, CA, Morgan Kaufmann, 1990.Google Scholar
- Y.Bengio, S. Bengio. Learning a synaptic learning rule,Tech. Rep. 751, Dép. d'Informatique et de Recherche Opérationnelle, Université de Montréal, Montreal (QC) Canada, 1990.Google Scholar
- V. N. Vapnik.Estimation of dependencies based on empirical data, Springer-Verlag, New-York, NY, USA, 1982.Google Scholar
- J. Holland.Adaptation in Natural and Artificial Systems, University of Michigan Press, 1975.Google Scholar
- S. Kirkpatrick, C. D. Gelatt, M.P. Vecchi, Optimization by simulated annealing,Science, no. 220, pp. 671–680, 1983.Google Scholar
- J. R. Koza.Genetic programming: on the programming of computers by means of natural selection, Bradford Book, MIT Press, Cambridge, MA, USA, 1992.Google Scholar
- S. Bengio, Y. Bengio, J. Cloutier, J. Gecsei. On the optimization of a synaptic learning rule, inConference on Optimality in Biological and Artificial Networks, Dallas, USA, 1992.Google Scholar