ESTIMATION OF RELATIONS BETWEEN LOGISTICS PERFORMANCE COMPONENTS BY ARTIFICIAL NEURAL NETWORKS

Abstract

Countries will be able to determine the degree of logistics performance input components to provide the logistics output components, and countries will be able to reveal which logistics input components should be carried out to provide output components. This situation enables countries to make sense of their contributions to the global economy and global trade. In this context, the relationship structure between logistics performance input components and logistics performance output components over the values of Logistic Performance Index (LPI) components of the countries in 2018, 2016, 2014, 2012 and 2010 LPI reports has been predicted with Artificial Neural Networks. According to the findings, the predictive values of contribution to the affective structure of the input components were listed as infrastructure, customs and service quality, and the predictive values of contribution to the influence structure of the output components as timeliness, international shipments and tracking and trace. Subsequently, it was found that the independent variables included in the analysis have quite high predictive power for dependent variables. Finally, the significance and normalized significance estimation significance values of the independent variables are listed as infrastructure, service quality and customs. In the study, it was observed that the contribution predictions of customs and service quality logistic input components to the structure of affecting the logistic output components were less than the infrastructure component. Therefore, it has been concluded that countries should carry out activities that will provide logistics performance output components of service quality and customs components in order to increase their own logistics performance and to contribute more to the global economy and trade.