If you originally shared by all neurons at excitatory and inhibitory, it significantly limits the ability of a neural network with a given structure. Why not use a one type of neurons. Are there any known electrical models of neurons, providing both intensification and weakening of synaptic connections?
Article Neural coding in graphs of bidirectional associative memories
It's not the neurons that are excitatory or inhibitory-but the synapses. There are analog models, however, like the Hodgkin-Huxley model, where the synapse doesn't just have a sign, that means that it's excitatory or inhibitory, but, also, a value. And there, are, of course, many other related developments, like integrate-fire models and so on.
Stam Nicolis, thanks for the reply! Note the following. Depending on the states of the interacting neurons same synapses may be involved in excitation and inhibition. Separation of neurons and synapses in the excitatory and inhibitory elements, is applicable only to solving a particular task.
One should be quite careful to distinguish artificial neural networks and models for neurobiology. Cf. here, http://bluebrain.epfl.ch/, for instance, for information on an effort to describe the latter.
- Badges
- Science topic
- Similar topics
- Biological Science
- Bioecology
- Systems Ecology
- Landscape Ecology
- Connectivity