Today, we met with the developers and were able to ask questions. The software uses nine different forecasting models. At any given time, the computer is able to determine which of the forecasting models provides the greatest accuracy at that time. It does this by finding which model has the smallest mean squared error. This was an exciting time, because I was able to see some of the concepts that I teach in statistics in action.
They use a forecasting model that has 39 coefficients and they have over 40,000 inventory items.
No comments:
Post a Comment