摘要
将L-M算法与填充函数法相结合,提出一种训练前向网络的混合型全局优化GOBP(G lobalOptim izationBP)算法。L-M算法的收敛速度快,利用它先得到一个局部极小点,然后利用填充函数算法跳出局部最小,得到一个更低的局部极小点,重复计算即可得到全局最优点。经实验验证,该算法收敛速度很快,避免了局部收敛,而且性能稳定。
Proposes GOBP(Global Optimization BP) neural networks in which combines the filled function method and Levenberg-Marquardt algorithm for training feed forward neural networks. With the L-M algorithm whose astringency is good, it can find one of local minimal points quickly. Afterwards, the filled function method will be used to find the point that is lower than the minimal point previously found. By repeating these processes, a global minimal point can be obtained at last. Practical examples indicate that the method has a higher accuracy in astringency and works well in avoiding sticking in local minima.
出处
《计算机应用研究》
CSCD
北大核心
2006年第2期211-212,255,共3页
Application Research of Computers
基金
山东省自然科学基金重大项目(Z2004G02)
山东省中青年科学家奖励基金资助项目(03BS003)
关键词
L-M算法
填充函数
全局优化
BP网络
L-M Algorithm
Fitted Function
Global Optimization
BP Neural Networks