Applied Online Algorithms with Heterogeneous Predictors
Applied Online Algorithms with Heterogeneous Predictors
复制标题
DOI：
--
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Jessica Maghakian;Russell Lee;M. Hajiesmaili;Jian Li;R. Sitaraman;Zhenhu Liu
中科院分区：
文献类型：
--
作者：
Jessica Maghakian;Russell Lee;M. Hajiesmaili;Jian Li;R. Sitaraman;Zhenhu Liu
在许多应用领域，机器学习（ML）模型与决策的集成受到黑箱模型较差的可解释性和理论保证的阻碍。虽然预测算法的新兴领域提供了一种利用ML的方法，同时享受最坏情况的保证，但现有的工作通常假设只有一个预测器。我们演示了如何通过有意使用不同数量的预测器来更有效地利用历史数据集和应用领域知识。通过利用我们的预测器的异质性，我们能够比预测不可知的方法获得更好的性能、可解释性和计算效率。理论结果辅以生产数据的大规模经验评估，证明了我们的方法在大型分布式计算系统中出现的优化问题上的成功。
For many application domains, the integration of machine learning (ML) models into decision making is hindered by the poor explainability and theoretical guarantees of black box models. Although the emerging area of algorithms with predictions offers a way to leverage ML while enjoying worst-case guarantees, existing work usually assumes access to only one predictor. We demonstrate how to more effectively utilize historical datasets and application domain knowledge by intentionally using predictors of different quantities. By leveraging the heterogeneity in our predictors, we are able to achieve improved performance, explainability, and computational efficiency over predictor-agnostic methods. Theoretical results are supplemented by large-scale empirical evaluations with production data demonstrating the success of our methods on optimization problems occurring in large distributed computing systems.