Research On Forecasting The Cost Of Residential Construction Based On SVM And LSSVM

To forecast the Cost of rough high level and non-assembled residential building project quickly and accurately before the design and the relevant information rarely, in analysis of the strengths and weaknesses in previous studies related theories and methods, We learned that based on machine learning theory to forecast the Cost of building Project is an important and feasible way, which is accurately and fast in today’s construction project cost forecasting.In this study, we will analysis deep into some studies on the cost of construction engineering based on machine learning theory. Residential construction project cost forecasting model based on principal component analysis and support vector machine is put forward. Then, based on the factors analysis in the cost of construction project and view a large number of construction project cost forecasting about index selection based on machine learning theory in old times, we selected some indicators based on machine learning theory to forecast the cost of building project, through the development of cost forecasting index system selection and adjustment rules. Realized the rules of model chosen which is based on the machine learning theory to forecast the cost of building Project, through selection of some sample about the cost of building Project, at this time we need to set some rules and standards about model chosen which is based on the machine learning theory to forecast the cost of building Project.This study will selects some indicators via selection rules to predict the cost of Construction and selects some sample via some rules and standards about sample project chosen, and then, combine the indicators and sample projects into the model we chose via rules of model chosen. At last we conducted an analysis based on real-life case scenarios to forecast the cost of constructions by the right indicators and the right model.