Introduction


SAS High-Performance Forecasting

SAS High-Performance Forecasting (HPF) software provides a system of SAS procedures for large-scale automatic forecasting in business, government, and academic applications. Major uses of High-Performance Forecasting procedures include: forecasting, forecast scoring, market response modeling, and time series data mining.

The software includes the following automatic forecasting process:

  • accumulates the time-stamped data to form a fixed-interval time series

  • diagnoses the time series using time series analysis techniques

  • creates a list of candidate model specifications based on the diagnostics

  • fits each candidate model specification to the time series

  • generates forecasts for each candidate fitted model

  • selects the most appropriate model specification based on either in-sample or holdout-sample evaluation using a model selection criterion

  • refits the selected model specification to the entire range of the time series

  • creates a forecast score from the selected fitted model

  • generate forecasts from the forecast score

  • evaluates the forecast using in-sample analysis

  • provides for out-of-sample forecast performance analysis

  • performs top-down, middle-out, or bottom-up reconciliations of forecasts in the hierarchy