This work aims to increase the service level and to reduce the inventory costs by combining the
forecast and inventory model into one consistent forecast-based inventory model. This new model
is based on the prediction of the future probability distribution by assuming an integer-valued
autoregressive process as demand process. The developed algorithms can be used to identify
estimate and predict the demand as well as optimize the inventory decision of intermittent
demand series. In an extensive simulation study the new model is compared with a wide range of
conventional forecast inventory model combinations. By using the consistent approach the mean
inventory level is lowered whereas the service level is increased. Additionally a modern
multi-criteria inventory classification scheme is presented to distinguish different demand
series clusters.
