Generalization and differentiation of affective associative memory circuit based on memristive neural network with emotion transfer.

Abstract

Most existing research on affective associative memory neural network circuits has predominantly concentrated on reinforcement and extinction, with insufficient attention given to the integration of emotion transfer alongside the principles of generalization and differentiation. To enhance the functionality of associative memory, this paper introduces a memristor-based affective associative memory neural network that incorporates generalization and differentiation under emotion transfer. The circuit implementation is carried out using PSPICE. The capabilities of learning, forgetting, emotion transfer, as well as generalization and differentiation under emotion transfer, are effectively realized and validated through simulation results. By examining generalization and differentiation under emotion transfer, this study presents an innovative approach to the field of affective associative memory neural networks based on memristors, particularly in relation to complex biological behaviors.