Poisson Regression in one Covariate on Massive Data

Torsten Reuter, Rainer Schwabe

The goal of subsampling is to select an informative subset of all observations, when using the full data for statistical analysis is not viable. We construct locally $D$-optimal…