Το έργο με τίτλο Exploring uncertainty, sensitivity and robust solutions in mathematical programming through Bayesian analysis από τον/τους δημιουργό/ούς Tsionas, Mike, Philippas Dionisis, Zopounidis Konstantinos διατίθεται με την άδεια Creative Commons Αναφορά Δημιουργού 4.0 Διεθνές
Βιβλιογραφική Αναφορά
M. G. Tsionas, D. Philippas and C. Zopounidis, “Exploring uncertainty, sensitivity and robust solutions in mathematical programming through Bayesian analysis,” Comput. Econ., vol. 62, no. 1, pp. 205–227, June 2023, doi: 10.1007/s10614-022-10277-z.
https://doi.org/10.1007/s10614-022-10277-z
The paper examines the effect of uncertainty on the solution of mathematical programming problems, using Bayesian techniques. We show that the statistical inference of the unknown parameter lies in the solution vector itself. Uncertainty in the data is modeled using sampling models induced by constraints. In this context, the objective is used as prior, and the posterior is efficiently applied via Monte Carlo methods. The proposed techniques provide a new benchmark for robust solutions that are designed without solving mathematical programming problems. We illustrate the benefits of a problem with known solutions and their properties, while discussing the empirical aspects in a real-world portfolio selection problem.