摘要
讨论了神经网络与纠错编码之间的关系.指出非线性分组码软判决最小距离译码等价于求神经网络能量函数最大值.
The relationship between the neural network and error--correcting code is discussed. It is given that the minimum distance soft--decision decodes of nonlinear block codes is equated to maximum valum of neural network energy function.
关键词
神经网络
能量函数
非线性分组码
neural networks
energy functions
nonlinear block codes
soft - decision decocling