摘要
研究了利用误差反向传播人工神经网络(BP网络)的多指标综合评价问题,并以建设工程项目招标评标为背景,建立了相应的多指标综合评价BP模型。引入附加动量法和变步长算法对BP网络算法进行改造,大大提高了标准BP神经网络收敛速度与学习训练速度,为解决多指标综合评价问题提供了一条有效途径。
The multiple attribute synthetical evaluation problems with error back propagation training artificial neural network (BP neural network) is researched. To comment the mark is the background to invite tenders by the construction project item, BP' s models of the multiple attribute syntherical evaluation have been established . The BP neural network algorithm has been transformed by lead into the additional momentum law and turn length of stride algorithm, the convergence speed and study training speed has been greatly raised, and it provides an effective approach to solving the problem of multiple attribute synthetical evaluation.
出处
《郑州轻工业学院学报（自然科学版）》
CAS
2003年第2期11-14,共4页
Journal of Zhengzhou University of Light Industry:Natural Science
关键词
人工神经网络
多指标综合评价
建设工程项目
招标
评标
artificial neural network
multiple attribute synthetical evaluation
the construction project calls for bid and evaluates the tender
