摘要
基于BP神经网络模型,提出保证成本估算精度的神经网络分析方法;并收集20个住宅工程实例,其中4个作为检测实例。实验结果表明根据工程前期阶段的基本信息推算得到的成本精度,可以满足实际工程成本估算的需要。
Puts forward a neural-network method to improve the cost estimation accuracy. 20 residential projects are gathered and four of them are taken in detection, the results show that the accuracy of cost is qualified for cost estimation in engineering operation, which is calculated by that basic prophase information.
出处
《建筑经济》
北大核心
2012年第12期89-91,共3页
Construction Economy
关键词
成本估算
神经网络
实例检测
BP算法
cost estimation
Neural Network
test sample
Back Propagation
