Research On Rationality Evaluation Of Construction Cost Based On Data Mining

The rationality evaluation of construction project cost runs through the whole process of project management.The purpose of evaluation is to discover and control the error of cost compilation in time,which is of great significance to improve the investment efficiency of construction project.However,in the current practice,the cost evaluation method is cumbersome,repetitive and technical lag,which leads to the cost evaluation work can not be effectively implemented.In order to improve the efficiency and technical level of cost evaluation,ensure the accuracy of cost compilation and the effectiveness of cost management,it is necessary to process the accumulated cost data,apply data mining technology to cost evaluation,and realize the transformation of cost evaluation mode.Firstly,the paper analyzes the problems existing in the application and evaluation of construction project cost data,puts forward the reasonable evaluation method of construction project cost based on data mining,and uses data processing technology to standardize,structurally integrate and process the cost data according to the difficult problems in the application of cost data;Then,the types and functions of construction projects are preliminarily classified.The unit project takes the architectural decoration project as an example,and 15 characteristic indexes are selected.The divisional project takes the pile foundation project as an example,and10 characteristic indexes are selected.Fuzzy c-means clustering(FCM)is used to identify the samples of similar unit projects and divisional projects as the training samples of the prediction model;Secondly,through the comparative analysis of prediction methods,the least squares support vector machine(LSSVM)is selected to establish the cost prediction model,and the reasonable cost intervals of unit project and divisional project are calculated respectively to evaluation the compilation cost;Finally,we use the historical cost data of 295 residential projects to evaluation the case project budget,and analyze and verify the feasibility and rationality of the construction project cost evaluation based on FCM and LSSVM model.FCM clustering can quickly and effectively identify samples of similar projects as evaluation models,which improves the subjectivity and randomness of sample selection in previous studies;Compared with support vector machine(SVM),LSSVM has better prediction performance;The proposed evaluation model can improve the efficiency and reliability of construction cost evaluation.The research conclusion not only helps to enrich the theory and method of project cost evaluation,but also improves the limitations of previous research focusing on cost index evaluation,and improves the application research of data mining in the basis and method of cost evaluation;It can also help the cost Auditors to complete the cost evaluation work efficiently,contribute to the rational planning and implementation of investment plans,so as to improve the investment efficiency of construction projects,and provide some reference for the application of big data in the informatization reform of project cost in the construction industry.