Research On Subcontractor Selection Based On Machine Learning Under Big Data

In the era of big data,data has become a means of production and a scarce asset.Any industry and field will produce valuable data,and the statistics,analysis,mining and utilization of these data will create unexpected value and wealth.With the increasing amount of data generated in the whole process of project subcontracting,we need to improve the ability of data analysis and data mining to find the huge value behind it,so as to improve the ability of subcontractor evaluation and selection.Through data analysis,machine learning technology,and analysis of business data,find valuable information,combining with the support of relevant theories and algorithm logic,to achieve scientific selection of subcontractors under big data.This paper focus on the selection of engineering subcontractors,and chooses a representative construction group in Nanjing as the reference object.After fully analyzing the business situation and data feature of the selection of engineering subcontractors,it determines a new idea of subcontractor selection,and constructs a subcontractor selection scheme based on machine learning under big data.First of all,based on the analysis of the data assets of a general contractor enterprise,the thesis constructs a sub contractor selection theme data warehouse,stores the scattered,heterogeneous,and different specifications of data in the enterprise in a unified way according to different dimensions to achieve the unified management of the sub contractor data,and improves the quality and use efficiency of the data related to the sub contractor management.Secondly,for the problem of subcontractor selection,we propose a subcontractor selection scheme based on machine learning under big data:(1)Based on the understanding of subcontractor selection and the consideration of the actual situation of the enterprise,we choose to extract the characteristics of subcontractors from three aspects: willingness to cooperate,project delivery ability,and company strength,and choose K-means algorithm to achieve clustering analysis of subcontractor data,Then,based on the analysis of the evaluation characteristics of subcontractors,the evaluation grade is divided to form a clear classification of subcontractors to support the decision-making of subcontractor selection.(2)The model of subcontractor selection is established through BP neural network,and the effective mechanism of subcontractor selection is established.The model is packaged into a software service module by software engineering method,and the optimal subcontractor can be selected only by entering new subcontractor information.The research on subcontractor selection based on machine learning and big data in order to solving the problems of difficult data acquisition,complex process and long selection cycle when general contracting enterprises select subcontractors,meeting the needs of enterprises for rapid and accurate decision-making,improving work efficiency,reducing work costs,and improving the objectivity of subcontractor selection process.This scheme enriches the theory,scheme and implementation method of engineering subcontractor selection based on machine learning and big data,give an new way for engineering subcontractor selection under big data.The whole idea has certain reference significance for the selection of other suppliers in the engineering industry and the application method of big data in the engineering industry.