BP算法在网络性能评估中的应用  

The Application of BP Algorithm to the Function Evaluation of Network

作　　者：原慧琴[1] 杨文伟[2] 

机构地区：[1]广东工程职业技术学院,广州510520 [2]广东省考试中心,广州510063

出　　处：《现代计算机》2006年第11期34-36,共3页Modern Computer

摘　　要：本文利用神经网络BP算法建立网络性能评估的数学模型,采用各性能指标作为其输入,网络性能作为输出,基于最小二乘思想,采用梯度搜索技术,以期使网络的实际输出值与期望输出值的误差均方值最小。经实验证明,该数学模型具有较好的辨识精度。This paper puts forward a mathematics model. It applies the BP algorithm of neural network to the evaluation of network performace. It defines each performance index as its input and network perfonnace as its output. It is based on the principle of minimum two multiply and the gradient searching technology to minumize the error square value between the actual output and the expected output. Proved by experiment, this model has a fine recognizability precision.

关 键 词：网络性能 性能指标 神经网络 BP算法 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]