1 paper
I. V. Grossu, C. I. Ciuluvica
Inspired by the importance of both communication and feedback on errors in human learning, our main goal was to implement a similar mechanism in supervised learning of artificial n…