“基因算法”在人工神经元网络中的应用

在使用著名的ＢＰ算法来训练一个多层毅向神经网络时，首先要确定这个神经元网络的主结构以及网络的各种初始化参数，但拓扑结构和各种参数的确定到目前为止还无确定的规则可寻。本研究将基因算法引入人工神经元网络。通过它来自动确定神经元网络的拓扑结构以及具体参数，并给出了基因算法在神经元网络应用中的具体实现。

Application of Genetic Algorithms to Artificial Neural Networks

When implementing a multi-layer neural networks using standard backpropagation, the topology architectures and various parameter attings of neural networks are judged. Untill now, these decisions are normally achieved by rules of thumb. This paper introduces genetic algorithms into neural networks and presents a compelete solution about how to tune neural networks with genetic algorithms.