Application of economic model optimization algorithm for international trade based on big data technology

Abstract

The advent of the era of big data has posed greater challenges to the economic model of international trade and opportunities for international trade. In this paper, from the application of big data in international trade, we study the SVM-based big data clustering analysis method for international trade and propose to optimize the objective function in the iterative process by using the individual fitness function of the chicken flock optimization algorithm for the problem that SVM is easy to fall into the optimal local solution. Then, the steps of international trade big data analysis by constructing CSO-SVM are used to forecast international port throughput, international trade exports, and imports. The average absolute error rate of the trained CSO-SVM model for the prediction of international port throughput decreased to 6.19%, and the accuracy of the prediction of international trade export volume and export price reached 86.86% and 87.83%, respectively, and the accuracy of the prediction of international trade import demand and import price reached 86.71% and 87.07%, respectively. During the test, CSO-SVM has higher prediction performance than other models. The analysis based on big data can better quantify the challenges faced by international trade, optimize international logistics scheduling, and improve import and export planning.