Scholarlink

Memristors can simulate biomimetic synapses and the observed biological neural activity of the neural system, such as conditioned reflex, memory, and forgetting. The paper designs a memristor-bridge synaptic model and establishes a neural network based on this to simulate the impact of external stimuli on human emotions in the Albert experiment. Compared with single memristive neural networks, the proposed network can achieve the function of associative storage memory, reduce energy consumption, and accomplish the generalization process of associative memory, which realistically reflects the characteristics of human memory. The learning and generalization implemented in the associative memory verify the feasibility and effectiveness of the network circuit design.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Memristive Neural Network for Learning and Generalization Based on Albert Associative Memory

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Mi Zhang,
  • Hui Chang,
  • Xintong Yue,
  • Xinzhe Zhang,
  • Baoxing Han

摘要

Memristors can simulate biomimetic synapses and the observed biological neural activity of the neural system, such as conditioned reflex, memory, and forgetting. The paper designs a memristor-bridge synaptic model and establishes a neural network based on this to simulate the impact of external stimuli on human emotions in the Albert experiment. Compared with single memristive neural networks, the proposed network can achieve the function of associative storage memory, reduce energy consumption, and accomplish the generalization process of associative memory, which realistically reflects the characteristics of human memory. The learning and generalization implemented in the associative memory verify the feasibility and effectiveness of the network circuit design.