摘要
神经网络是人工智能应用的重要研究领域,因其出色的高度非线性映射能力、自组织和适应能力、记忆联想能力。已经成为机器学习研究的热点,本文详细讨论了多层神经网络的理论依据,泛化能力问题,总结神经网络结构优化与泛化能力的研究现状,最后提出了神经网络的未来发展。
The neural network is an applied important research in Artificial Intelligence, because of its outstanding nonlinear mapping, self-organized, paralleled, adaptation, and so on. This makes it an important research in machine learning. The paper has discussed the development of muhilayer feed-forward neural networks, the theory foundation, the generalization ability, and surveyed the latest research on the generalization ability and structural optimization of Neural Networks. At last, it shows clearly the future and the development of the Neural Network.
出处
《柳州师专学报》
2006年第3期88-95,共8页
Journal of Liuzhou Teachers College
基金
广西教育厅资助项目(200508234)
关键词
多层前向神经网络
泛化能力
结构优化
multilayer feed-forward neural networks
generalization ability
structural optimization
