AUTOMATIC DETECTION OF SYSTEMATIC SALES PATTERNS USING AUTOCORRELATION TECHNIQUE

摘要

A method, based on autocorrelation techniques, for measuring the relative significance of the systematic versus random components of product sales data. The results of this determination can be used to improve product demand forecast and product seasonal profile determinations. When a product's sales variation is primarily due to systematic patterns, the accuracy of demand predictions and forecasts can be improved by understanding and modeling the underlying pattern. On the other hand, when variations in sales are merely random, these variations can be discounted when determining demand forecasts or product seasonal profiles.