Forecasting methods can help companies to, among other things: plan and estimate the values of the investments to be made, the inventories to be created, the service capacity needed to provide a service, the size of the production, etc .; verify the effect of the entry of competitors; find out where it is necessary to make a sales effort and have different schedules of promotions and discounts. As a result of the forecasting process, overestimated estimates (forecast above the real) entail the so-called excess cost, which can manifest itself in the form of, for example, a higher fixed cost, an unnecessary cost of inventory, obsolescence, perishability, or an expense advertising. On the other hand, underestimated estimates (forecast below the real) cause the so-called stockout cost, which can take the form of, for example, loss of the contribution margin of a product or service, waiting orders in the production queue or negative consequences of a demand not answered. Therefore, the ideal is to make forecasts that are as close as possible to reality, that is, to create models that have the least possible prediction errors. With the exponentially increasing amount of information available to enterprises to use, the challenge of forecasting methods becomes even greater and more important, both computationally and analytically. This special issue aims to contribute to the explanation of how Data Science can help Forecasting Methods within this current Big Data scenario.

Aims and Scope:

1. Forecasting Methods
2. Data Science
3. Big Data
4. Forecasting models
5. Forescasting errors
6. Models improvement