Data-driven Prediction of Relevant Scenarios for Robust Combinatorial Optimization
Optimization and Control
2022-12-26 v2 Machine Learning
Abstract
We study iterative methods for (two-stage) robust combinatorial optimization problems with discrete uncertainty. We propose a machine-learning-based heuristic to determine starting scenarios that provide strong lower bounds. To this end, we design dimension-independent features and train a Random Forest Classifier on small-dimensional instances. Experiments show that our method improves the solution process for larger instances than contained in the training set and also provides a feature importance-score which gives insights into the role of scenario properties.
Cite
@article{arxiv.2203.16642,
title = {Data-driven Prediction of Relevant Scenarios for Robust Combinatorial Optimization},
author = {Marc Goerigk and Jannis Kurtz},
journal= {arXiv preprint arXiv:2203.16642},
year = {2022}
}