摘要
复数权神经元是由于引入多阈值逻辑而使得权值为复数的神经元,被认为具有更强的性能。该文根据该神经元数学模型,结合POST代数系统中多值逻辑取值及运算的概念,实现了基于单个复数权神经元的多值整形运算,设计的多值逻辑整形运算具有稳健性能,所得结果表明用复数权神经元实现多值逻辑的有效性和可行性,说明了其强大的信息处理能力。
Neurons with complex-valued weights have stronger capability because of its multi-valued threshold logic.In the paper,a multi-valued reshaper is implemented according to both the neuron's mathematics model and POST algebra system and its operation.More over the designed reshaper has robustness feature.So the result is believed that the validity and possibility with such kind of neurons for multi-valued logic as well as its strong capability for information processing.
出处
《杭州电子科技大学学报(自然科学版)》
2008年第1期20-23,共4页
Journal of Hangzhou Dianzi University:Natural Sciences
基金
杭州电子科技大学科学研究基金资助项目(KYF051507001)
关键词
复数权值
多值神经元
多值逻辑
稳健神经元
complex-valued weights
multi-valued neurons
multi-valued logic
robust neurons