摘要
神经网络集成是机器学习和神经计算重要研究领域,通过训练有限个神经网络个体,并将其结论进行适当的合成,可以极大提高学习系统的泛化能力,已经成为一种有广阔应用前景的工程化神经计算技术。本文讨论了神经网络集成的理论方法及其研究进展,并对该领域进一步研究的问题进行了探讨。
The Neural Network ensemble is an applied important research in machine learning and Neuro-computing, because it can significantly improve the generalization ability of neural network through ensembling a number of neural networks, i. e. training many neural networks and then combining their prediction. It has the broad application prospect since this technology behaves remarkably well based system. Recently it has become a very hot topic research. The paper discusses the development, theory foundation and at last, it shows clearly the future and the development of the Neural Network ensemble .
出处
《柳州师专学报》
2007年第4期118-122,共5页
Journal of Liuzhou Teachers College
基金
广西教育厅项目(200508234)
关键词
神经网络
神经网络集成
泛化误差
neural network
neural network ensemble
generation error