|
  |
|
|
Krasnow Institute > Monday Seminars > Abstracts Artificial Neural Networks Based on the Neurobiology of Associative Learning: Derivation and Performance. Tom Vogl Recently, some fundamental biochemical and biophysical changes at the subcellular level consequent to associative learning have been revealed in a marine mollusc and in mammals. Based on these findings, it has been possible to construct an artificial neural network that utilizes non-Hebbian learning rules and that exhibits useful properties including: monotonic convergence, rapid learning, self-organization, and weak dependence on global parameters. The neurobiology underlying this network, its architecture, and its performance on real-life tasks is presented.
The Krasnow Institute for Advanced Study |
|
