Identyfikacja obserwacji odstających w szeregach czasowych na przykładzie branży FMCG
Identification of outliers in the time series on an example of the FMCG industry

Summary/Abstract: The aim of the article is to verify and analyse selected methods for the identification of outliers in the FMCG industry (fast-moving consumer goods). Real tests were used to carry out the sales data of the American chain of Walmart stores, made available by the company on the Internet. The analysis has been made using classic methods for identifying outliers: three sigma method and quartile method, as well as an advanced method based on the auto regression model proposed in the 1993 work of Chen and Liu. Article begins with a short introduction and characteristics of the FMCG industry. Next, the definitions, types of outliers and methods for identification are presented. The paper ends with the presentation of research results for two units using all three methods.

• Details
• Contents