|
生物物理学报 1991
ON THE CONDITIONAL ASSOCIATIVE NEURAL NETWORK
|
Abstract:
A new neural network model, the conditional associative neural network, was proposed in this paper.The model consists of a conventional associative memory network and a presynapstic inhibitation network. The preliminary analysis and computer simulation show that the model has some capabilities that conventional associative memory model do not have, such as it can response to same input with different outputs under different conditions, and response to different inputs with same output under different conditions. The model will help us to understand some information processing in the brain. Furthermore, a possible neural realization of the model was Proposed.