一种用于计算机模拟的神经元模型
A Neuron Model for Computer Simulation of Neural Nets
-
摘要: 本文提出了一种单个神经元的模型,它能较好地解决计算机模拟语言所要求的通用性和 精确性之间的关系问题,通过引入S型曲线来描述突触上的记忆值,能给短时记忆和长时记 忆一个统一的刻画,并在突触水平上模拟了习惯化、敏感化、条件化和容易化等四种基本学习 形式.Abstract: In this paper, we propose a mathematical model of a single neuron, which possesses qualities of both generality and accuracy required by a computer simulation language. By introducing S-shape curve to describe the memory process of a memory synapse, long-term memory and short-term memory can be handled consistently, and four basic learning forms, i.e., habituation sensitization, conditioning and facilitation, can be modeled at the level of synapse.
-
点击查看大图
计量
- 文章访问数: 1435
- HTML全文浏览量: 112
- PDF下载量: 1066
- 被引次数: 0
下载:
