Robust optimization has been established as a leading methodology to approach decision problems under uncertainty. To derive a robust optimization model, a central ingredient is to identify a suitable model for uncertainty, which is called the uncertainty set, containing all scenarios against which we wish to protect. An ongoing challenge in the recent literature is to derive uncertainty sets from given historical data. In this paper we use an unsupervised deep learning method to construct non-convex uncertainty sets from data, which have a more complex structure than the typically considered sets. We prove that most of the classical uncertainty classes are special cases of our derived sets and that optimizing over it is strongly NP-hard. Nevertheless we show that the trained neural networks can be integrated into a robust optimization model by formulating the adversarial problem as a convex quadratic mixed-integer program. This allows us to derive robust solutions through an iterative scenario generation process. We prove that our class of uncertainty sets contains In extensive computational experiments, we compare this approach to a similar approach, which derives uncertainty sets by kernel-based support vector clustering. We find that uncertainty sets derived by the unsupervised deep learning method can give a better description of data, leading to robust solutions that often outperform the comparison method both with respect to objective value and feasibility.
翻译:强力优化已被确立为处理不确定性中决策问题的主导方法。 要形成一个稳健优化模式, 核心要素是确定一种适合不确定性的模式, 称为不确定性集, 包含我们希望保护的所有情景。 最近文献中的一项持续挑战是从给定的历史数据中获取不确定性组。 在本文件中, 我们使用一种未经监督的深层次学习方法, 从数据中构建非混凝土不确定性组, 这些数据组的结构比通常考虑的组群复杂得多。 我们证明, 古典不确定性类大多是我们衍生的数据集的特殊案例, 并且对之进行优化是很强的NP- 硬性。 然而, 我们表明, 受过训练的神经网络可以通过将对抗性问题发展成一个正统的优化模型, 将之作为二次二次二次二次二次方位方位方位混合整数程序。 这使我们能够通过反复的假想生成过程获得稳健的解决方案。 我们证明, 我们的不确定性组组组别包含广泛的计算实验, 我们把这个方法与类似的方法进行比较, 这种方法通过以内核支持的矢量组合产生不确定性组。 我们发现, 由未经监督的深层次学习方法产生的不确定性组分化的不确定性组, 能够更准确地描述数据的可行性方法, 更精确地描述数据。