Cooperative Artificial Neural Architecture
Cooperative Maximum-Likelihood Hebbian Learning
(CMLHL) [1] is based on the standard Maximum Likelihood
Network [13][16][12] but this time it has a lateral connection
(which acts after the feedforward but before the feedback)
derived from the Rectified Gaussian Distribution [18][1][19]
and using the cooperative distribution. The final architecture
is described as follows. There is a Feedforward step (1), then
there is a lateral activation passing: