@svetlyak40wt @spacebat Sorry for the late answer - my private life is absorbing most of my bandwidth in this period (divorce):
as far as I can see, my concept shares something with the regulatory feedback ANN model, but the layout of the network itself is modified by the feedback.
as far as I can see, my concept shares something with the regulatory feedback ANN model, but the layout of the network itself is modified by the feedback.
@svetlyak40wt @spacebat Think of it as a directed cyclic graph, with an input layer, an output layer and an arbitrary number of intermediate nodes; plus, it can transmit information and modify its layout (the connections, not the nodes) according to a feedback mechanism, ...
@svetlyak40wt @spacebat ... working its way unsupervised to a stable configuration that produces an optimal feedback.
@svetlyak40wt @spacebat The initial goal is not, of course, efficiency or practical applications, but producing a behavior that resembles biological learning. Then we could start exploring efficiency and applications.
@threadreaderapp: please unroll
@threadreaderapp: please unroll
• • •
Missing some Tweet in this thread? You can try to
force a refresh
