Adaptive estimation of daily demands with complex calendar effects fdr freight transportation

We address the problem of forecasting spatial activities on a daily basis t hat are subject to the types of multiple, complex calendar effects that ari se in many applications. Our problem is motivated by applications where we generally need to produce thousands, and frequently tens of thousands, of m odels, as arises in the prediction of daily origin-destination freight flow s. Exponential smoothing-based models are the simplest to implement, but st andard methods can handle only simple seasonal patterns. We propose a class of exponential smoothing-based methods that handle multiple calendar effec ts. These methods are much easier to implement and apply than more sophisti cated ARIMA-based methods. We show that our techniques actually outperform ARIMA-based methods in terms of forecast error, indicating that our simplic ity does not involve any loss in accuracy. (C) 2000 Elsevier Science Ltd. A ll rights reserved.