Integer Programming Formulations Applied to the Optimal Sample Allocation Problem

This article proposes two integer programming formulations, which were applied
to solve  the problem of optimal allocation  in stratified sampling. In one of the versions
From this problem, we seek to minimize a weighted sum of relative variances associated with m
research variables, a sample size n was set and that will be allocated to L population strata.
The two formulations were applied to a set of 21 populations, diversified as
'the origin, size, values of m and L population strata, contemplating four different scenarios
as to the value of the sample size n. From the results obtained and from a set of < /> analyses, it was possible to evaluate the efficacy and viability of the two formulations, compared to a formulation
of literature.