Analysis of the BP neural network comprehensive competitiveness evaluation model for the development evaluation of B2B E-commerce enterprises

Abstract

ABSTRACT In today’s steady development of e-commerce, improving enterprise competitiveness is a major challenge that e-commerce enterprises need to face. There are problems with incomplete and inaccurate evaluation systems in commonly used competitiveness evaluation systems. To address these concerns, the study establishes a model for evaluating business-to-business relationships by analyzing indicators of competitiveness. It also examines the algorithm process and neural network structure and incorporates them into the evaluation model. The experiment shows that through training the back propagation neural network, it is discovered that the mean square error of the samples is 9.9869e-06 when the training frequency reaches 3200, resulting in the completion of convergence of the back-propagation neural network. The utilization of back propagation neural networks to enhance the evaluation model of enterprise competitiveness has yielded positive outcomes. Suggestions have been proposed to improve the competitiveness of diverse e-commerce enterprises. This study can furnish enterprise managers with decision support, enhancing enterprise competitiveness and enabling sustainable development.