Project Overview:
We are looking for help in optimizing the configuration of the Demantra statistical engine. This is a process that consists of: Analyzing and segmenting the data (i.e., seasonality, intermittency, trending) Determining the appropriate level to forecast at ( "forecast tree")Creating profiles based on those segmentsAssigning models to those profilesAdjusting the parameters for those profiles and modelsFine-tuning the algorithms, considering seasonal trends, product demand variability, significance, and local causal factors. This tends to be fairly interactive - such as identifying combinations with suboptimal forecasts, understand the reasons and determine corrective actions.
