| China’s construction industry is changing with each passing day along with social development,but credit problems are also becoming increasingly prominent among market players,which need to be dealt with urgently.At present,the credit evaluation system of related enterprises in the construction market has been gradually established,but the study of the credit system of professionals is still relatively lacking,and it is difficult to establish a sound system of credit evaluation for practitioners in the construction industry.At the same time,as a project cost practitioner,his untrustworthy behavior not only hinders the construction of corporate credit and the healthy operation of the construction market,but also makes the cost work lose the role of dynamic control of project costs,which has a negative impact on project benefits.Therefore,the credit management of engineering cost practitioners has become one of the current key concerns.This thesis builds a credit evaluation system for engineering cost practitioners based on BP neural network.Firstly,through the study of relevant literature and national policies,laws and regulations,the credit evaluation indicators of engineering cost practitioners are extracted and summarized,and then the questionnaire survey and cluster analysis method are combined to optimize the evaluation index system.Finally,a credit evaluation index system of construction cost practitioners including 26 three-level indicators is constructed.Secondly,this thesis builds the BP neural network structure,that is,sets the appropriate learning parameters,and collects and processes data samples.On the one hand,the input layer data is parameterized,and on the other hand,the expected value of each sample is determined by combining the expert scoring method for the BP neural network simulation training.Provide preparation.Finally,this thesis uses MATLAB software to conduct simulation experiments,and uses 3-fold cross-validation method to expand the data sample,and finally obtain 150 prediction results.By comparing the expected output and the predicted results,the accuracy rate of the credit evaluation model for engineering cost practitioners in this thesis is as high as 96.00%,and the deviation is in the range of acceptance,which proves the evaluation model has a good prediction effect.Therefore,the use of BP neural network for credit evaluation of engineering cost practitioners not only avoids the problems of index weight assignment,complex evaluation process and low efficiency in traditional evaluation methods,but also improves the objectivity and scientificity of evaluation results and maintains certain accuracy.,which provides a new way for the credit management of construction cost practitioners in the construction industry.In addition,this thesis provides corresponding suggestions for the application of the credit system for construction cost practitioners from four aspects:information data quality assurance and sharing mechanism,supervision mechanism,reward and punishment mechanism,and credit education mechanism. |
