Research On Vendor Selection Of Government Procurement Based On Deep Neural Network

With the development of society,government procurement is no longer just to meet the operational needs of government departments to purchase goods or services,it has been entitled with more functions.It is a barometer reflecting the relevant policies of the government,a scale for allocating social resources,and a mirror reflecting the openness,fairness and integrity of the government.The objectives and functions of government procurement are becoming increasingly diversified.Suppliers are linked to government demand and market supply.The goal and effect of government procurement largely depends on the comprehensive level of government procurement suppliers.Combining with the characteristics and principles of government procurement,the selection of suppliers has become a field that scholars pay attention to.The research on this issue has important theoretical value and practical significance.This dissertation not only analyses the development status and trend of government procurement,but also studies the evaluation methods of government procurement suppliers in depth.At present,the index setting of supplier evaluation methods are not good enough to justify it's systematic logic and cutting-edge characteristic.On the other hand,current evaluation system only focus the product itself,and it fails to combine with the current trend of product service.Due to the strong subjectivity of the evaluation method or the insufficient complexity and practicality of the algorithm,the traditional supplier selection method is no longer applicable when encountering with multi-objective decision-making.In recent years,the rise of information processing methods based on artificial neural network(ANN)makes it possible to make more accurate and complex multi-objective decision-making using mathematical models and computer software.In this dissertation,a deep neural network method is proposed as a supplier evaluation method.By using the prism model and analyzing the stakeholders of suppliers,the evaluation system of suppliers is constructed from the needs,contributions,strategies,processes and capabilities of stakeholders.According to the index system,the method of deep neural network model is applied to the evaluation of government procurement suppliers,and the evaluation model of government procurement suppliers based on deep neural network is constructed.Finally,through empirical analysis,the application steps of the deep neural network evaluation model are realized by using MATLAB.The validity of the evaluation model is verified,which proves that the method is effective and feasible,and some suggestions are put forward to apply the method to practice.This dissertation constructs an evaluation index system and method model for supplier selection and evaluation of government procurement,which has the following significance: First,it is a vital supplement to the theory of supplier evaluation,enriches and improves the current theory of supplier research,and has certain theoretical value.Second,the evaluation index system of government procurement suppliers is comprehensive,the model is easy to operate,and has strong feasibility.It provides a reference method for the selection of government procurement suppliers.