We introduce a novel measure for quantifying the error in input predictions. The error is based on a minimum-cost hyperedge cover in a suitably defined hypergraph and provides a general template which we apply to online graph problems. The measure captures errors due to absent predicted requests as well as unpredicted actual requests; hence, predicted and actual inputs can be of arbitrary size. We achieve refined performance guarantees for previously studied network design problems in the online-list model, such as Steiner tree and facility location. Further, we initiate the study of learning-augmented algorithms for online routing problems, such as the traveling salesperson problem and dial-a-ride problem, where (transportation) requests arrive over time (online-time model). We provide a general algorithmic framework and we give error-dependent performance bounds that improve upon known worst-case barriers, when given accurate predictions, at the cost of slightly increased worst-case bounds when given predictions of arbitrary quality.
翻译:我们引入了一种量化输入预测错误的新措施。 错误基于在适当定义的高光谱中以最低成本高端覆盖为基础, 提供了我们适用于在线图形问题的一般模板。 该措施捕捉了因缺少预测请求以及未预知的实际请求而产生的错误; 因此, 预测和实际投入可能具有任意性; 我们为先前研究过的网上列表模型网络设计问题( 如施泰纳树和设施位置)实现了完善的绩效保障。 此外, 我们开始研究在线路径问题( 如旅行销售人员问题和拨号问题)的学习强化算法, 这些问题( 运输) 要求到达时( 在线时间模型 ) 。 我们提供了一个一般算法框架, 并给出基于错误的性能约束, 在给出准确的预测时, 改进已知的最坏的屏障, 以给出任意质量预测时略微增加的最坏的框框的成本 。