Big Data are generally huge quantities of digital information accrued automatically and/or merged from several sources and rarely result from properly planned population surveys. A Big Dataset is herein conceived as a collection of information concerning a nite population. Since the anal- ysis of an entire Big Dataset can require enormous computational eort, we suggest selecting a sample of observations and using this sampling information to achieve the inferential goal. Instead of the design-based survey sampling approach (which relates to the estimation of summary nite population measures, such as means, totals, proportions) we con- sider the model-based sampling approach, which involves inference about parameters of a super-population model. This model is assumed to have generated the nite population values, i.e. the Big Dataset. Given a super-population model we can apply the theory of optimal design to draw a sample from the Big Dataset which contains the majority of in- formation about the unknown parameters of interest. In addition, since a Big Dataset might provide poor information despite its size, from the def- inition of eciency of a design we suggest a device to measure the quality of the Big Data.
Deldossi, L., Tommasi, C., Optimal Design of Experiments and Model-based survey sampling in Big-Data, in Programme and Abstracts, 19th Annual ENBIS Conference, Budapest, 2-4 september 2019, (Budapest (Ungheria), 02-04 September 2019), Jens Bischoff, Agnes Backhausz, Rossella Berni, Sonja Kuhnt, Lluıs Marco- Almagro, Antonio Pievatolo, Marco P. Seabra dos Reis and Murat Caner Testik, Budapest, Hungary 2019:2019 37-37 [http://hdl.handle.net/10807/147206]
Optimal Design of Experiments and Model-based survey sampling in Big-Data
Deldossi, Laura
Primo
;
2019
Abstract
Big Data are generally huge quantities of digital information accrued automatically and/or merged from several sources and rarely result from properly planned population surveys. A Big Dataset is herein conceived as a collection of information concerning a nite population. Since the anal- ysis of an entire Big Dataset can require enormous computational eort, we suggest selecting a sample of observations and using this sampling information to achieve the inferential goal. Instead of the design-based survey sampling approach (which relates to the estimation of summary nite population measures, such as means, totals, proportions) we con- sider the model-based sampling approach, which involves inference about parameters of a super-population model. This model is assumed to have generated the nite population values, i.e. the Big Dataset. Given a super-population model we can apply the theory of optimal design to draw a sample from the Big Dataset which contains the majority of in- formation about the unknown parameters of interest. In addition, since a Big Dataset might provide poor information despite its size, from the def- inition of eciency of a design we suggest a device to measure the quality of the Big Data.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.