Predictive analytics in manufacturing uses models extracted from historical order and production related records to support the optimization of manufacturing processes. For instance, forecasting the overall amount of incoming orders is crucial for a) properly scheduling and sizing raw material and part orders and b) configuring processes as e.g. final assembly or testing. Loosely based on a specific subproblem in one of our ongoing R&D projects, this howto showcases the process of building an order amount predictor in R.
