Autotune is a tool to evaluate different settings for forecast model parameters. The objective is to find the optimal parameter settings that generate the most accurate forecasts.

The total number of forecasts to evaluate increases with every setting. The tool makes one forecast for every combination of settings for the chosen historic periods and evaluates them with measured data to find the best setting. The chosen setting can then be applied for future forecasts as the primary model, a partially weighted model or as a non-weighted model.

Autotune can take a lot of time for a computer if there are many chosen settings and/or long periods to test, which renders a lot of calculations. Therefore it can be good to select settings and periods with care. Furthermore it is possible to divide the calculations to several computers, which speeds up the process.