TY - GEN
T1 - A Learning-Based Mathematical Programming Formulation for the Automatic Configuration of Optimization Solvers
AU - Iommazzo, Gabriele
AU - D’Ambrosio, Claudia
AU - Frangioni, Antonio
AU - Liberti, Leo
N1 - Publisher Copyright:
© 2020, Springer Nature Switzerland AG.
PY - 2020/1/1
Y1 - 2020/1/1
N2 - We propose a methodology, based on machine learning and optimization, for selecting a solver configuration for a given instance. First, we employ a set of solved instances and configurations in order to learn a performance function of the solver. Secondly, we formulate a mixed-integer nonlinear program where the objective/constraints explicitly encode the learnt information, and which we solve, upon the arrival of an unknown instance, to find the best solver configuration for that instance, based on the performance function. The main novelty of our approach lies in the fact that the configuration set search problem is formulated as a mathematical program, which allows us to a) enforce hard dependence and compatibility constraints on the configurations, and b) solve it efficiently with off-the-shelf optimization tools.
AB - We propose a methodology, based on machine learning and optimization, for selecting a solver configuration for a given instance. First, we employ a set of solved instances and configurations in order to learn a performance function of the solver. Secondly, we formulate a mixed-integer nonlinear program where the objective/constraints explicitly encode the learnt information, and which we solve, upon the arrival of an unknown instance, to find the best solver configuration for that instance, based on the performance function. The main novelty of our approach lies in the fact that the configuration set search problem is formulated as a mathematical program, which allows us to a) enforce hard dependence and compatibility constraints on the configurations, and b) solve it efficiently with off-the-shelf optimization tools.
KW - Automatic algorithm configuration
KW - Hydro unit committment
KW - Machine learning
KW - Mathematical programming
KW - Optimization solver configuration
UR - https://www.scopus.com/pages/publications/85101257387
U2 - 10.1007/978-3-030-64583-0_61
DO - 10.1007/978-3-030-64583-0_61
M3 - Conference contribution
AN - SCOPUS:85101257387
SN - 9783030645823
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 700
EP - 712
BT - Machine Learning, Optimization, and Data Science - 6th International Conference, LOD 2020, Revised Selected Papers
A2 - Nicosia, Giuseppe
A2 - Ojha, Varun
A2 - La Malfa, Emanuele
A2 - Jansen, Giorgio
A2 - Sciacca, Vincenzo
A2 - Pardalos, Panos
A2 - Giuffrida, Giovanni
A2 - Umeton, Renato
PB - Springer Science and Business Media Deutschland GmbH
T2 - 6th International Conference on Machine Learning, Optimization, and Data Science, LOD 2020
Y2 - 19 July 2020 through 23 July 2020
ER -