Critical Echo State Networks that Anticipate Input using Morphable Transfer Functions
This addresses a problem for researchers in reservoir computing and neuromorphic engineering by proposing a novel approach, though it appears incremental as it builds on existing echo state network concepts.
The paper tackles the problem of improving echo state networks by introducing truly critical networks with morphable transfer functions that anticipate inputs, resulting in deviations being forgotten slowly in a power law fashion, as analyzed numerically in a one-neuron model.
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.