|
Abstract:
|
An important task for any large-scale business is to prepare forecasts of business metrics, such as revenue, cost, and event occurrences, at different time horizons (e.g. weekly or quarterly intervals). Often these business organizations are structured in a hierarchical manner by line of business, division, geography, product line or a combination thereof. In many situations projections for these business metrics may have been obtained independently and for each level of the hierarchy. The problem with forecasts produced in this way is that there is no guarantee that forecasts are aggregate consistent according to the hierarchical structure of the business, while remaining as accurate as possible. In addition, it is often important for the organization to achieve accurate forecasts at certain levels of the hierarchy according to the needs of users. We develop a novel Bayesian approach to hierarchical forecasting that provides an organization with optimal forecasts that reflect their preferred levels of accuracy while maintaining the proper additive structure of the business.
|
