Preparing data series

Before sales can be forecasted, it is extremely important to adjust individual data series (i.e., sales of individual items) for algorithmic forecasting.

First and foremost, it means to clean the data by removing extreme sales and adjust the sales for a stockout (the product would be sold, if there were enough stock).

In this section, we deal with various analyzes of the data series: they will – if the right mathematical algorithm is applied – help us to choose a correct method of sales forecast. One of the most important parts in this section is an analysis of seasonality of sales.

It is also in this section, where we work with new products; to that end, we prepare short sales series to predict (manage) their supply.

Modules:

• Module: Managing new products and sell-outs
• Module: Computing seasonality
• Module: Computing historical stock levels
• Module: Analysis of stockouts
• Module: Analysis of extremes
• Module: Trimming irrelevant history
• Holidays
• Module: Cancellations adjustments