Logistics Modeling and Forecasting with Regression

In this chapter, the method of multiple regression is introduced for describing the functional relationship among several variables, as well as for predicting the values of a variable from the values of a group of variables. The tools for model fitting, model validation, and prediction are presented, while emphasis is given on understanding the types of data that can be analyzed via regression. More specifically, the method of least squares is discussed. Regression analysis is proposed due to its simplicity and wide applicability. Modeling outsourcing or demand forecasting can both be achieved by regression analysis, providing useful information for logistics service providers or 3PL companies. Hauling freight data collected from a logistics company based in Ohio were utilized to demonstrate the applicability of regression analysis and its usefulness for logistics service providers, 3PL companies, and transportation companies. Finally, limitations, solutions, and alternative strategies are discussed.