Dipartimento di Scienze Statistiche

  The Cube Sampling method is the most used technique for selecting balanced samples. It is composed of two phases: the flight phase, which can be performed efficiently, and the landing phase, which may require the solution of a Linear Programming problem with an exponential number of variables. In this paper some variants of the landing phase of the Cube method are proposed with the aim of reducing the dimension of the Linear Programming problem to be solved and then improving on the total execution time. The algorithmic efficiency of the proposed techniques is thoroughly evaluated via an experimental study.