The paper deals with the problem of temporal aggregation in time series forecasting. The question is, is it better to use aggregate data? For example, is it better to use weekly figures rather than daily?
We discuss the perfromance of temporal aggregation approaches. We assess the forecast combination of temporal aggregation and investigate how temporal aggregation affects time series features. We examine the association between the original time series features and the performance of approaches using machine learning
This seminar provides an overview of the research dealing with aggregation and hierarchical forecasting