摘要
针对人工神经网络技术在水文研究中出现的过拟合现象,从训练样本的构成入手,提出了两种新的训练策略,即择优检验和加权检验。前者通过计算比较,选择最优的训练样本;后者对不同训练样本的评定结果进行加权,以得到最终的结果。应用实例表明,此两种方法能有效地减轻神经网络中过拟合的缺点。
Aiming at the problem of over training in hydrologic research using artificial neural networks (ANNs)and considering the composition of training samples, two new learning strategies were presented, i. e, optimizeddetecting method and weighted detecting method. The former chooses the optimal training data after comparingthe validation sets. The latter endues the different sets with different weights in order to get a final validationoutcome. The application of the two new methods at Zipingpu Hydrographic Station on Minjiang River upstreamindicates that they can resolve the problem of over-training effectively.
出处
《长江科学院院报》
CSCD
北大核心
2002年第3期59-61,共3页
Journal of Changjiang River Scientific Research Institute
基金
国家自然科学基金重大项目(50099620)
关键词
神经网络
加权检验
洪水预测
neural networks
optimized dotecting method
weighed detecting method
flood prediction
