Applications of large deviations to optimal experimental designs
Résumé
We give a large deviations principle for the least-squares estimator in a linear model. Next, we apply the large deviations result to find optimal experimental designs.