Abstract: | The objective of this research was to estimate logistic expenditure for the global economy. An Artificial Neural Network was used to generate national estimates. The estimation set constitutes 24 countries representing approximately 75% of the global Gross Domestic Product. The model utilizes variables that capture economic activity, transportation activity, income level, country size, and geographic location. This research updates previous estimates and represents continuous improvement of the estimation method, including new input variables, better information quality, mathematical model refinement, and methods of model validation. |