The paper investigates a new type of truly critical echo state networks where individual transfer functions for every neuron can be modified to anticipate the expected next input. Deviations from expected input are only forgotten slowly in power law fashion. The paper outlines the theory, numerically analyzes a one neuron model network and finally discusses technical and also biological implications of this type of approach.
@article{arxiv.1606.03674,
title = {Critical Echo State Networks that Anticipate Input using Morphable Transfer Functions},
author = {Norbert Michael Mayer},
journal= {arXiv preprint arXiv:1606.03674},
year = {2017}
}
Comments
14th International Symposium on Neural Networks (ISNN), Sapporo, Hakodate, Japan, June 21st - 26th 2017