Neural Circuit For Self-regulated Attentional Learning In Selective Attention Adaptive Resonance Theory (saart) Neural Networks

Abstract

This paper presents a novel neural mechanism of self regulated attentional learning for real-time competitive neural networks that are also capable of selective attention. The neural circuit, currently aimed at SAART neural networks, detects whether the input is familiar or novel and then uses the generated signals to modulate synaptic signal transfer at various stages.