摘要
针对标准BP算法存在的缺陷,本文给出了基于MATLAB语言的BP神经网络几种改进的算法.阐述了各种BP算法的优化技术原理、优缺点,并就它们的训练速度和内存消耗情况作了比较.建议在多数BP神经网络训练时,先尝试使用Levenberg-Marquardt算法,其次是BFGS算法或共轭梯度法以及RPROP算法.
To eliminate the shortcoming of standard backpropagation algorithm, some modified BP algorithms in the MATLAB's neural networks toolbox are given in the paper. These high performance algorithms are discussed in the optimization techniques and compared with speed and memory. For most situations, the Levenberg-Marquardt algorithm is advised to use first, then the BFGS algorithm, or one of the conjugate gradient methods, or RPROP algorithm is considered.
出处
《科技通报》
北大核心
2003年第2期130-135,共6页
Bulletin of Science and Technology