摘要
为实现建设工程造价的快速和准确预测,此文提出基于模糊最小二乘支持向量机的建设工程造价预测方法。该方法可较好解决小样本预测问题,适合于当前工程造价样本数据量少的现状。通过隶属度函数对样本进行模糊化和加权,实现对历史数据和相似数据的优化选择,提高了预测准确性。将标准SVM的二次规划问题转化为线性方程组求解,提高了预测速度。通过对某市地铁建设中区间隧道延米造价估算实例的计算,验证了所提出预测方法的有效性。
To realize the fast and accurate prediction of construction engineering cost, the fuzzy least square support vector machine forecasting method of the project cost is put forward. This method can solve prediction problem of small sample well, which is suitable for the small sample data problem in the current project cost. Through fuzzifing and weighting by the membership functions, the historical data and the optimal selection of similar data is realized, which improves the prediction accuracy. The standard SVM quadratic programming problem is transformed into solving linear equations to improve the prediction speed. Taking the subway tunnels Yanmi cost estimate as an example, the validity of the proposed prediction method is veryfied.
出处
《铁路工程造价管理》
2012年第3期1-4,18,共5页
Railway Engineering Cost Management
关键词
工程造价
支持向量机
隶属度
预测
engineering cost
support vector machine （SVM）
membership
forecast
