In various approaches to learning, notably in domain adaptation, active learning, learning under covariate shift, semi-supervised learning, learning with concept drift, and the like, one often wants to compare a baseline classifier to one or more advanced (or at least different) strategies. In this chapter, we basically argue that if such classifiers, in their respective training phases, optimize a so-called surrogate loss that it may also be valuable to compare the behavior of this loss on the test set, next to the regular classification error rate. It can provide us with an additional view on the classifiers' relative performances that error rates cannot capture. As an example, limited but convincing empirical results demonstrates that we may be able to find semi-supervised learning strategies that can guarantee performance improvements with increasing numbers of unlabeled data in terms of log-likelihood. In contrast, the latter may be impossible to guarantee for the classification error rate.
研究了基于最小化凸代理的机器学习方法的一致性性质。我们扩展了Osokin et al.（2017）最近的框架，用于对不一致代理的情况下的一致性属性进行定量分析。我们的关键技术贡献在于一个新的下限的校准函数的二次代理，这是不平凡的（不总是零）不一致的情况下。新的界限允许量化设置的不一致性水平，并显示如何使用不一致的代理学习可以保证样本的复杂性和优化难度。我们将我们的理论应用于两个具体案例：具有树结构损失的多类分类和具有平均精度损失的排名。结果表明，近似计算的权衡所造成的不一致的代理人和他们的潜在利益。
Quantifying the Uncertainty in Model Parameters Using Gaussian Process-Based Markov Chain Monte Carlo: An Application to Cardiac Electrophysiological Models
Rochester Institute of Technology（罗切斯特理工学院）；Dalhousie University（达尔豪斯大学）
Goal-oriented learning of stochastic differential equations using error bounds on path-space observables
Center for Computational Science & Engineering, Massachusetts Institute of Technology（麻省理工学院计算科学与工程中心）；Department of Mathematics, University of Potsdam（波茨坦大学数学系）
Scalability vs. Utility: Do We Have to Sacrifice One for the Other in Data Importance Quantification?
Virginia Tech（弗吉尼亚理工大学）；UIUC（伊利诺伊大学厄巴纳-香槟分校）；Shanhai Jiaotong University（上海交通大学）；UC Irvine（加州大学尔湾分校）；ETH Zurich（苏黎世联邦理工学院）；Lawrence Livermore National Laboratory（劳伦斯利弗莫尔国家实验室）；UC Berkeley（加州大学伯克利分校）
量化每个训练点对学习任务的重要性是机器学习中的一个基本问题，并且已经利用估计的重要性分数来指导一系列数据工作流，例如数据汇总和域自适应。一个简单的想法是使用每个训练点的留一误差来表示其重要性。最近的工作也提出了使用Shapley值，因为它定义了一个唯一的值分布方案，满足一组吸引人的属性。然而，计算Shapley值通常是昂贵的，这限制了它在大规模实际应用中的适用性。最近已经提出了多个算法来提高计算Shapley值的可扩展性，但这可能会影响它们在实际应用中的实用性。\textit{How well do existing data quantification methods perform on existing workflows? How do these methods compare with each other, empirically and theoretically? Must we sacrifice scalability for the utility in these workflows when using these methods?}在本文中，我们进行了一种新的理论分析比较不同的重要性量化方法的效用，并报告了广泛的实验研究现有的和拟议的工作流程，如噪声标签检测，水印去除，数据汇总，数据采集，和领域自适应。我们表明，Shapley值近似的基础上，$K$ …
Supervised Pretraining for Material Property Prediction
Lane Department of Computer Science and Electrical Engineering, West Virginia University（兰恩计算机科学与电气工程系，西弗吉尼亚大学）；Department of Physics and Astronomy, West Virginia University（物理与天文学系，西弗吉尼亚大学）
Quantifying Uncertainty in Space Debris Capture with Active Tether-Net Systems Caused by Noisy Observations
Department of Mechanical and Aerospace Engineering, AIAA Student Member（机械与航空航天工程系，AIAA学生会员）；Department of Mechanical and Aerospace Engineering, AIAA Member（机械与航空航天工程系，AIAA会员）；Department of Mechanical and Aerospace Engineering, AIAA Senior Member（机械与航空航天工程系，AIAA高级会员）
Wearable data from subjects playing Super Mario, sitting university exams, or performing physical exercise help detect acute mood episodes via self-supervised learning
School of Informatics, University of Edinburgh（爱丁堡大学信息学院）；Hospital Clínic de Barcelona, University of Barcelona（巴塞罗那大学巴塞罗那临床医院）；Institute of Psychiatry, Psychology and Neuroscience, King’s College London（伦敦国王学院精神病学、心理学与神经科学研究所）；Division of Psychiatry, University of Edinburgh（爱丁堡大学精神病学系）
