Forecasting in Database Systems

Time series forecasting is crucial in a number of domains such as production planning and energy load balancing. In these areas, forecasts are often required by non-expert users on large multi-dimensional data sets expecting short response times. However, as current traditional database systems support forecasting only in a limited and non-declarative way, it is performed outside the database system by specially trained experts. We introduce a novel approach that seamlessly integrates time series forecasting into an existing database management system. In contrast to flash-back queries that allow a view on the data in the past, we have developed a Flash-Forward Database System (FDB) that provides a view on the data in the future. It supports a new query typeÐa forecast queryÐ that enables forecasting of time series data for any user and is automatically processed by the core engine of an existing DBMS. We introduce various optimization techniques for three different types of forecast queries: ad-hoc queries, recurring queries, and continuous queries. All approaches intend to increase the efficiency of forecast queries while ensuring high forecast accuracy.