Port container throughput prediction method based on SSA-SVM

Abstract

To improve the prediction accuracy of the port cargo throughput and the applicability of the prediction model, and then provide data support for the port construction to meet the needs of port decision-making, take the monthly cargo throughput data of Shanghai Port from January 2009 to December 2022 as an example, use Pearson correlation analysis to screen 12 import and export impact factors. This paper improves the traditional SVM model, uses SSA (Sparrow Search Algorithm) to optimize the parameters c and g in SVM, and uses the model to predict. Compared with the model that uses the grid search algorithm to optimize the parameters of SVM, the model has a significant improvement in fitting and robustness, its predicted value is closer to the actual value, the prediction performance is better, and it can better reflect the actual state of the port.