Multi-Attribute Comprehensive Evaluation of Government Procurement Suppliers: A Biotechnological Framework

Abstract

The traditional methodologies for evaluating government procurement suppliers often falter under the complex demands of modern performance assessments. This paper introduces an innovative neural network-based decision model designed to enhance the robustness of multi-objective decision-making and address the complexities and uncertainties inherent in conventional neural network methods for supplier evaluation. The proposed model integrates a supplier evaluation decision framework with advanced mathematical techniques to optimize decision-making processes. To validate the accuracy and effectiveness of the model, we conducted a comparative analysis of the government procurement evaluation algorithm against existing methods. Experimental results demonstrated a significant improvement, with the Mean Squared Error (MSE) of the model decreasing from 0.7653 to 0.1289 through the implementation of a Particle Swarm Optimization-Backpropagation (PSO-BP) algorithm, similar in performance to the PSO-Adam algorithm. The findings confirm that our decision-making model offers a robust, objective, and effective approach for supplier evaluation in challenging operational environments. By minimizing subjective bias and simplifying the evaluation process, the model contributes to more efficient supply chain management and reduced operational costs. This approach is particularly pertinent to the biotechnology sector, where precise supplier evaluation can impact both production efficiency and compliance with stringent industry standards.