Contribution
Trieste - Italy
Adding a weak, stochastic, nonspecific reinforcement process to the clipped perception algorithm lets it efficiently solve the supervised learning problem in neuronal models with binary synapses
Speakers
- CARLO BALDASSI
Primary authors
- CARLO BALDASSI (ISI Institute for Scientific Interchange Foundation, Torino, Italy)
Organizers
Directors: N. Brunel, P. Del Giudice, S. Franz, S. Fusi, R. Zecchina