摘要
基于递阶结构的遗传算法可以同时对多层前向神经网络进行结构优化和权重求解。与基本的遗传算法相比,这种算法不仅在权重训练方面更加快速稳定,而且能在学习过程中确定网络的拓扑结构,具有较高的学习效率,而在遗传过程中采用自适应的交叉和变异概率能有效加快遗传速度和避免早熟现象的出现。
The genetic algorithm based hierarchical structure could optimize the structure and weights of neural network simultaneously. Compared with simple genetic algorithm, it has more efficiency in not only accelerating and stabilizing the weights training but also determining the network's structure. Adopting adaptive crossover and mutation probability could accelerate the genetic speed and avoid prematurity.
出处
《计算机工程与设计》
CSCD
北大核心
2005年第2期305-307,共3页
Computer Engineering and Design
基金
江苏省教育厅自然基金资助项目(03KJB510041)。
