Year
2018
Authors
ALFANDARI Laurent, ESPINOZA GARCIA J.-C.
Abstract
We propose a robust approach for solving Linear Programs the coefficients of which depend on uncertain data in a non-linear way, like NPVs with the discount rate in finance. We propose first a binary approximation of the non-linear functions, then a piece-wise linear approximation. In the second case, numerical experiments show that the robust solution remains feasible for more than 6000 instances of problems tested, including Capital Budgeting, despite the approximation of the non-linear functions.
ALFANDARI, L. et ESPINOZA GARCIA, J.C. (2018). Robust Optimization for Non-Linear Impact of Data Variation. Computers & Operations Research, 99, pp. 38-47.