论文部分内容阅读
针对工程项目成本预估问题,提出基于LS-SVM的工程项目成本预估模型。首先构建包括定性指标和定量指标的成本预估指标体系,并对定性预估指标量化赋值;然后采用模糊模式识别方法筛选出用于模型构建的输入样本工程项目集,并采用主成分分析法提取预估指标主成分;最后基于最小二乘支持向量机回归建立预估模型。经实例验证,所建立的工程项目成本预估模型具有一定的实用性和可靠性。
Aiming at the project cost estimation problem, a project cost estimation model based on LS-SVM is proposed. Firstly, the cost forecasting index system including qualitative index and quantitative index is constructed, and the qualitative index is quantified and assigned. Then the input sample project set for model construction is screened by fuzzy pattern recognition method and extracted by principal component analysis Predict the main components of the index; Finally, based on least squares support vector machine regression to establish the predictive model. The example validation shows that the established project cost forecast model has certain practicality and reliability.