Neural Diversity Machines
Although most Artificial Neural Networks (ANN) are universal function approximators,
if we constrain their sizes, many can be relatively ineffective for certain kind of problems.
The question that motivates this study is whether we can find an ANN that can be
effective for a broader set of problems even when it is relatively compact.
References
- Maul, T., Bargiela A., Ren, L.J., Cybernetics of Vision Systems: Towards an Understanding of Putative Functions of the Outer Retina, IEEE Transactions on Systems Man and Cybernetics, SMC-A, 41(3), 398-409, 2011, doi:10.1109/TSMCA.2010.2085432
- Maul, T.H., Bargiela, A., Ren, L.J. A Neuroalgorithmic Investigation of the Outer Retina. Proc. 24th European Conference on Modelling and Simulation, ECMS 2010, Kuala Lumpur, Malaysia, pp.337-343, June 2010
To download a PDF file of the poster please click on the image