摘要: 针对传统BP神经网络权值算法速度慢、易陷入局部极小等缺陷,在权值平衡算法的基础上,提出了一种激励函数参数可调的前馈神经网络,并给出了相应的权值和参数快速学习算法。该算法运用该文提出的非单调启发式模拟退火搜索法实现网络权值和参数的快速搜索。实验表明,该算法不仅能明显提高网络的学习速度,而且可较好地避免学习过程陷入局部极小点而导致学习失败。
关键词:
可调参数,
神经网络,
模拟退火法,
权值平衡,
快速学习算法
Abstract: Aiming at traditional BP neutral network weight algorithm’s defect such as slow convergence and easy to plunge into local extremum, an adjustable parameter feedforward neutral network and a fast algorithm to train it are proposed on the basis of weight balance algorithm. Based on the search technique of non-monotone heuristic simulated annealing proposed, the algorithm realizes fast search of network weight and parameter. Experiments show that this algorithm can speed up the learning process of network, and solve the problem of local extremum in learning process to a certain extend.
Key words:
Adjustable parameter,
Neural network,
Simulated annealing,
Weight balance,
Fast learning algorithm
中图分类号:
彭小奇;王 文;宋彦坡;张建智. 一种可调参数前馈神经网络的快速学习算法[J]. 计算机工程, 2007, 33(08): 187-189.
PENG Xiaoqi; WANG Wen; SONG Yanpo; ZHANG Jianzhi. Fast Learning Algorithm of Adjustable Parameter Feedforward[J]. Computer Engineering, 2007, 33(08): 187-189.