Memristor-based affective associative memory neural network circuit with emotional gradual processes

• Published in: Neural computing & applications Vol. 34; no. 16; pp. 13667 - 13682
• Main Authors: Liao, Meiling, Wang, Chunhua, Sun, Yichuang, Lin, Hairong, Xu, Cong
• Format: Journal Article
• Language: English
• Published: London Springer London 01.08.2022; Springer Nature B.V
• Subjects: Artificial Intelligence; Associative memory; Circuit design; Computational Biology/Bioinformatics; Computational Science and Engineering; Computer Science; Data Mining and Knowledge Discovery; Emotions; Image Processing and Computer Vision; Learning; Memristors; Neural networks; Neurons; Original Article; Probability and Statistics in Computer Science; Associative memory; Memristive neural network; Conditioning reflex; Circuit simulation; Affective model
• ISSN: 0941-0643; 1433-3058
• DOI: 10.1007/s00521-022-07170-z

In the existing affective associative memory neural network circuits, the change of emotions in the affective associative learning and forgetting processes is abrupt and the intensity of emotions is invariable. In fact, the transition from one emotion to another is a gradual process. In this paper, to realize the progressive changes of emotional intensity in the affective associative memory neural network, the gradual learning, gradual forgetting and gradual transferring processes of emotions are proposed and the memristor-based circuit of the affective associative memory neural network is designed. In the designed circuit, the firing frequency of output neurons is closely correlated with the intensity of emotions. The higher the firing frequency of output neurons, the stronger the emotional intensity. Based on the associative memory rule, the dynamical change of the synaptic weights leads to the gradual variation of the frequencies of output neurons. Thus, the function of variable emotional intensity can be realized and the gradual processes can be achieved. The PSPICE simulation results are given to verify that the proposed circuit could realize the affective learning, forgetting and transferring functions with gradual processes.