论文部分内容阅读
近年来,人们开发了许多的神经网络自动设计方法,其中进化算法与神经网络的结合最引人注目,形成一类进化人工神经网络。对用遗传算法实现神经网络的进化进行研究的时间最早,并得到广泛的评论,而最近越来越多的研究支持利用进化规划设计神经网络是一种更优的方法。该文讨论了这一领域的研究进展并阐明了进一步的研究方向。
In recent years, a lot of methods have been developed for the automatic design of neural networks. Among them, the combination of evolutionary algorithms and neural networks is the most conspicuous, forming an evolutionary artificial neural network. The earliest research on the evolution of neural networks by using genetic algorithm has been widely reviewed. However, more and more researches support the use of evolutionary programming to design neural networks, which is a better method. This article discusses the research progress in this area and clarifies further research directions.