欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
BP算法是目前应用极为广泛的神经网络算法，但它也存在一些不足。文中提出了一种使BP网络结构简化的改进的BP算法，它使得网络的速度得到了提高，减少了迭代次数。最后将传统的BP算法和改进的BP算法进行了比较，仿真结果表明改进的算法在学习次数和收敛速度上比传统的算法有很大的改善。
英文摘要：
BP is a neural network algorithm which is applied very widely, but it has some defects. So brings forward a method which can improve BP algorithm through simplifying BP＇ s structure on this paper. The method improves the training speed of the BP neural net. The method reduces the iterative times onBP neural net. In the end, compares the traditional algorithm with the improved algorithm on this paper. The results through simulation experinaent indicate that the improved algorithm has a large improvement on training times and convergence speed than the traditional algorithm.
同期刊论文项目
同项目期刊论文
期刊信息
位置：