An overview of the delelopment of approaches to construction cost and price forcasting since the 1950’s is given. First, second and third generation models can be identified, but they all have shortcomings. This pape...An overview of the delelopment of approaches to construction cost and price forcasting since the 1950’s is given. First, second and third generation models can be identified, but they all have shortcomings. This paper puts forward a new model, fuzzy calculation model, based on lots of data of the finished proiects. Through actual application, it is proved that the model is accurate and quick in calcalation of construction.展开更多
工程造价快速估算问题一直是工程管理领域的研究热点之一。为了研究建筑领域工程造价的快速估算方法,本文利用线性回归、SVM、K近邻、回归树、集成模型五类回归模型进行住宅成本预测。通过数值仿真模拟实验分析快速估算成本模型精度与速...工程造价快速估算问题一直是工程管理领域的研究热点之一。为了研究建筑领域工程造价的快速估算方法,本文利用线性回归、SVM、K近邻、回归树、集成模型五类回归模型进行住宅成本预测。通过数值仿真模拟实验分析快速估算成本模型精度与速度,并对模型影响因子进行分析。结果表明:集成模型下Extra Trees Regressor算法表现最优,运算时间上稍逊于无参数参与的K近邻回归算法。在贡献度上,优越功能的贡献度最高,层高的贡献度最小,其模型精度、速度满足实际工程成本预测的要求。展开更多
文摘An overview of the delelopment of approaches to construction cost and price forcasting since the 1950’s is given. First, second and third generation models can be identified, but they all have shortcomings. This paper puts forward a new model, fuzzy calculation model, based on lots of data of the finished proiects. Through actual application, it is proved that the model is accurate and quick in calcalation of construction.
文摘工程造价快速估算问题一直是工程管理领域的研究热点之一。为了研究建筑领域工程造价的快速估算方法,本文利用线性回归、SVM、K近邻、回归树、集成模型五类回归模型进行住宅成本预测。通过数值仿真模拟实验分析快速估算成本模型精度与速度,并对模型影响因子进行分析。结果表明:集成模型下Extra Trees Regressor算法表现最优,运算时间上稍逊于无参数参与的K近邻回归算法。在贡献度上,优越功能的贡献度最高,层高的贡献度最小,其模型精度、速度满足实际工程成本预测的要求。
