Jordan Constabileo

New Machine learning service is flawed. We have some products that are sold once a year or every other year. The model will extrapolate the demand as incremental and creates a forecast that is way too high. The old ML classic handled these scenarios perfectly. Now with the new model, we have to do too much analysis to find these incorrectly forecasted items. Why cant the new ML service behave in the same manner. Also, the Microsoft Demand Management App has the same flaw.