no code implementations • 9 Aug 2018 • Felipe Campelo, Fernanda Takahashi
In this work we present a methodology for defining the required sample sizes for designing experiments with desired statistical properties for the comparison of two methods on a given problem class.
