Econometric Count Data Forecasting and Data Mining (Cluster Analysis) Applied to Stochastic Demand in Truckload Routing
The challenge of truckload routing is increased in complexity by the introduction of stochastic demand. This paper examines this stochastic truckload demand using an econometric discrete choice model known as a Count Data Model, and applies it to the truckload routing problem (TRP) which is presented using a triplet formulation. We use actual truckload demand data and data from the bureau of transportation statistics to facilitate Poisson and negative binomial bases for count data regression. Regardless of the base distribution, two outcomes are produced from every regression run. These are the predicted count that represents the discrete truckload demand between every origin and destination and the likelihood that the count will occur. These two outcomes provide us with the ability to generate the traditional expected value of truckload demand as input to the TRP formulation. The inclusion of probabilities representing likelihoods results in a more robust technique for forecasting truckload demand.
Miori, Virginia M. "Econometric Count Data Forecasting and Data Mining (Cluster Analysis) Applied to Stochastic Demand in Truckload Routing." Advances in Business and Management Forecasting 6 (2009): 191-216. Print.
This document is currently not available here.