arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2009.14552math.OCcs.LGstat.ML
收藏

Wasserstein分布鲁棒逆多目标Optimization

Wasserstein Distributionally Robust Inverse Multiobjective Optimization

Amazon(亚马逊);University of Pittsburgh(匹兹堡大学)

Chaosheng Dong, Bo Zeng

摘要:

逆多目标优化提供了一个通用的框架,为无监督学习任务的推断参数的多目标决策问题(ESTA),根据一组观察到的决定,从人类专家。然而,这个框架的性能关键依赖于一个准确的决策,足够的高质量的决策,和一个参数空间,包含足够的信息的决策的可用性。为了避免假设条件、数据和参数空间的不确定性,本文研究了逆多目标优化的分布鲁棒方法。具体来说,我们利用Wasserstein度量来构建一个以这些决策的经验分布为中心的球。然后,我们制定了Wasserstein分布鲁棒逆多目标最优化问题(WRO-IMOP),最大限度地减少了最坏情况下的预期损失函数,其中最坏的情况下采取了所有的分布在Wasserstein球。我们证明了WRO-IMOP估计的超额风险具有次线性收敛速率。此外,我们提出了半无限的WRO IMOP的重新制定和开发的切割平面算法,收敛到一个近似解在有限的迭代。最后,我们证明了我们的方法的有效性的合成多目标二次规划和现实世界的投资组合最优化问题。

英文摘要:

Inverse multiobjective optimization provides a general framework for the unsupervised learning task of inferring parameters of a multiobjective decision making problem (DMP), based on a set of observed decisions from the human expert. However, the performance of this framework relies critically on the availability of an accurate DMP, sufficient decisions of high quality, and a parameter space that contains enough information about the DMP. To hedge against the uncertainties in the hypothetical DMP, the data, and the parameter space, we investigate in this paper the distributionally robust approach for inverse multiobjective optimization. Specifically, we leverage the Wasserstein metric to construct a ball centered at the empirical distribution of these decisions. We then formulate a Wasserstein distributionally robust inverse multiobjective optimization problem (WRO-IMOP) that minimizes a worst-case expected loss function, where the worst case is taken over all distributions in the Wasserstein ball. We show that the excess risk of the WRO-IMOP estimator has a sub-linear convergence rate. Furthermore, we propose the semi-infinite reformulations of the WRO-IMOP and develop a cutting-plane algorithm that converges to an approximate solution in finite iterations. Finally, we demonstrate the effectiveness of our method on both a synthetic multiobjective quadratic program and a real world portfolio optimization problem.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑