ASACT - Data preparation for forecasting: A method to substitute transaction data for unavailable product consumption data

The proposed ASACT (Aggregate, Smooth, Aggregate, Convert to Time-series) method applies aggregation and smoothing to transform delivery records into time-series data that creates a good approximation of actual consumption behavior; noise in the data is reduced while behavior patterns are preserved. Synthetic data is used to test the method against traditional methods, and real data is used to demonstrate the method's application in industry. Tests have shown that ASACT improves on the results produced by traditional methods.