摘要
BP算法是前馈神经网络训练中应用最多的算法 ,但其具有收敛慢和陷入局部极值的严重缺点。为了克服其缺点 ,近年来人们做了大量的工作。本文从改进的BP算法、智能优化算法、混合学习策略等方面评述了目前应用于前馈神经网络优化的算法 ,对各种算法的优缺点进行了分析 ,并给出克服现有算法缺陷的对策。
BP algorithm is the most popular training algorithm for feed forward neural network learning. But falling into local minimum and slow convergence are its drawbacks. Recent years, much work has been done to overcome these drawbacks. In this paper, improved BP algorithms, intelligent optimization algorithms and hybrid learning strategies have been surveyed. The advantages and disadvantages of each algorithm have been analyzed. Some countermeasures for overcoming these drawbacks have been presented.
出处
《机床与液压》
北大核心
2003年第5期29-32,共4页
Machine Tool & Hydraulics
基金
教育部高等学校骨干教师资助计划资助
关键词
前馈神经网络
BP算法
智能优化算法
混合学习策略
Feed forward neural network
BP algorithm
Intelligent optimization algorithm
Hybrid learning strategy
