|
创作新主题 |
社区所有版块导航
DATA
| docker Elasticsearch |
WEB开发
| linux MongoDB Redis DATABASE NGINX 其他Web框架 web工具 zookeeper tornado NoSql Bootstrap js peewee Git bottle IE MQ Jquery |
机器学习
| 机器学习算法 |
产品
| 短视频 |
印度
| 印度 |
一周十大热门主题
Update！H5支持摘要折叠，体验更佳！点击阅读原文访问arxivdaily.com，涵盖CS|物理|数学|经济|统计|金融|生物|电气领域，更有搜索、收藏等功能！
cs.LG 方向，今日共计64篇
Graph相关(图学习|图神经网络|图优化等)(3篇)
【1】 Graph Learning for Cognitive Digital Twins in Manufacturing Systems
标题：制造系统中认知数字孪生的图学习
链接：https://arxiv.org/abs/2109.08632
机构：Department of Electrical Engineering and Computer Science†, Department of Mechanical and Aerospace Engineering‡, University of California, Irvine, California, USA
摘要：未来的制造业需要将仿真平台和虚拟化与工业过程中的物理数据连接起来的复杂系统。数字双胞胎包括物理双胞胎、数字双胞胎以及两者之间的连接。使用数字孪生兄弟（尤其是在制造业）的好处很多，因为它们可以提高整个制造生命周期的效率。随着时间的推移，由于许多技术的发展，数字孪生兄弟的概念变得越来越复杂和强大。在本文中，我们详细介绍了认知数字孪生兄弟作为数字孪生兄弟的下一个发展阶段，这将有助于实现工业4.0的愿景。认知数字双胞胎将使企业能够创造性地、有效地、高效地利用从现有制造系统经验中获得的隐性知识。它们还支持更自主的决策和控制，同时提高整个企业的性能（规模）。本文介绍了图形学习作为一个潜在的途径，使认知功能的制造数字双胞胎。提出了一种在制造业产品设计阶段利用图形学习实现认知数字孪生的新方法。
摘要：Future manufacturing requires complex systems that connect simulation platforms and virtualization with physical data from industrial processes. Digital twins incorporate a physical twin, a digital twin, and the connection between the two. Benefits of using digital twins, especially in manufacturing, are abundant as they can increase efficiency across an entire manufacturing life-cycle. The digital twin concept has become increasingly sophisticated and capable over time, enabled by rises in many technologies. In this paper, we detail the cognitive digital twin as the next stage of advancement of a digital twin that will help realize the vision of Industry 4.0. Cognitive digital twins will allow enterprises to creatively, effectively, and efficiently exploit implicit knowledge drawn from the experience of existing manufacturing systems. They also enable more autonomous decisions and control, while improving the performance across the enterprise (at scale). This paper presents graph learning as one potential pathway towards enabling cognitive functionalities in manufacturing digital twins. A novel approach to realize cognitive digital twins in the product design stage of manufacturing that utilizes graph learning is presented.
【2】 Dynamic Spatiotemporal Graph Convolutional Neural Networks for Traffic Data Imputation with Complex Missing Patterns
标题：动态时空图卷积神经网络在复杂缺失模式交通数据填充中的应用
链接：https://arxiv.org/abs/2109.08357
机构：Department of Urban Planning and Design, The University of Hong Kong, Hong Kong, China, Department of Civil Engineering, McGill University, Montreal, QC, Canada
摘要：数据缺失是智能交通系统中交通数据采集不可避免且普遍存在的问题。尽管对交通数据插补进行了广泛的研究，但仍存在两个局限性需要解决：第一，现有方法无法捕捉交通数据中复杂的时空相关性，尤其是随时间演化的动态空间相关性；其次，以前的研究主要集中在随机缺失模式上，而对其他更复杂的缺失场景讨论较少。为了填补这些研究空白，我们提出了一种称为动态时空图卷积神经网络（DSTGCN）的新的深度学习框架来填充缺失的交通数据。该模型将递归结构与基于图的卷积相结合来建模时空依赖关系。此外，我们还引入了一种图结构估计技术，从实时交通信息和道路网络结构中对动态空间相关性进行建模。在两个公共交通速度数据集上进行了大量实验，将我们提出的模型与最先进的深度学习方法在四种缺失模式下进行了比较。结果表明，我们提出的模型在各种缺失场景下都优于现有的深度学习模型，并且图结构估计技术有助于提高模型的性能。我们进一步将我们提出的模型与张量因子分解模型进行比较，发现在不同的训练方案和数据可用性下，不同模型族之间存在不同的行为。
摘要：Missing data is an inevitable and ubiquitous problem for traffic data collection in intelligent transportation systems. Despite extensive research regarding traffic data imputation, there still exist two limitations to be addressed: first, existing approaches fail to capture the complex spatiotemporal dependencies in traffic data, especially the dynamic spatial dependencies evolving with time; second, prior studies mainly focus on randomly missing patterns while other more complex missing scenarios are less discussed. To fill these research gaps, we propose a novel deep learning framework called Dynamic Spatiotemporal Graph Convolutional Neural Networks (DSTGCN) to impute missing traffic data. The model combines the recurrent architecture with graph-based convolutions to model the spatiotemporal dependencies. Moreover, we introduce a graph structure estimation technique to model the dynamic spatial dependencies from real-time traffic information and road network structure. Extensive experiments based on two public traffic speed datasets are conducted to compare our proposed model with state-of-the-art deep learning approaches in four types of missing patterns. The results show that our proposed model outperforms existing deep learning models in all kinds of missing scenarios and the graph structure estimation technique contributes to the model performance. We further compare our proposed model with a tensor factorization model and find distinct behaviors across different model families under different training schemes and data availability.
【3】 Learning Sparse Graph with Minimax Concave Penalty under Gaussian Markov Random Fields
标题：高斯马尔可夫随机场下具有极小极大凹罚的稀疏图学习
链接：https://arxiv.org/abs/2109.08666
机构：ORTEGA., Department of Electronics and Electrical Engineering, Keio University, Kanagawa ,-, Japan, Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, California, CA , USA
备注：11 pages, 7 figures
摘要：提出了一种从数据中学习稀疏图的凸分析框架。虽然我们的问题公式的灵感来自于使用所谓的组合图拉普拉斯框架的图形套索的扩展，但关键区别在于使用非凸替代$\ellu_1$范数，以获得具有更好解释性的图形。具体来说，我们使用弱凸极大极小凹惩罚（$\ell_1$范数和Huber函数之间的差异），已知该惩罚可以产生稀疏解，其估计偏差低于回归问题的$\ell_1$。在我们的框架中，图拉普拉斯在优化中被对应于其上三角部分的向量的线性变换所取代。通过基于Moreau分解的重新表述，我们证明了通过在代价函数中引入二次函数，可以保证整体凸性。利用原始-对偶分裂方法可以有效地解决这一问题，并给出了可证明收敛的容许条件。数值算例表明，在合理的CPU时间下，该方法明显优于现有的图学习方法。
摘要：This paper presents a convex-analytic framework to learn sparse graphs from data. While our problem formulation is inspired by an extension of the graphical lasso using the so-called combinatorial graph Laplacian framework, a key difference is the use of a nonconvex alternative to the $\ell_1$ norm to attain graphs with better interpretability. Specifically, we use the weakly-convex minimax concave penalty (the difference between the $\ell_1$ norm and the Huber function) which is known to yield sparse solutions with lower estimation bias than $\ell_1$ for regression problems. In our framework, the graph Laplacian is replaced in the optimization by a linear transform of the vector corresponding to its upper triangular part. Via a reformulation relying on Moreau's decomposition, we show that overall convexity is guaranteed by introducing a quadratic function to our cost function. The problem can be solved efficiently by the primal-dual splitting method, of which the admissible conditions for provable convergence are presented. Numerical examples show that the proposed method significantly outperforms the existing graph learning methods with reasonable CPU time.
Transformer(3篇)
【1】 Primer: Searching for Efficient Transformers for Language Modeling
标题：Primer：寻找高效的语言建模转换器
链接：https://arxiv.org/abs/2109.08668
机构：Google Research, Brain Team
备注："Primer: Searching for Efficient Transformers for Language Modeling" initial preprint. 35 pages
摘要 ：大型转换器模型是自然语言处理最新进展的核心。然而，这些模型的训练和推理成本增长迅速，变得昂贵得令人望而却步。在这里，我们的目标是通过寻找更高效的Transformer来降低Transformer的成本。与以前的方法相比，我们的搜索是在较低的级别上执行的，在定义Transformer TensorFlow程序的原语上执行。我们确定了一个名为Primer的体系结构，它比原始Transformer和其他用于自回归语言建模的变体具有更小的训练成本。Primer的改进主要归因于两个简单的修改：平方ReLU激活和在自我注意的每个Q、K和V投影后添加深度卷积层。实验表明，Primer相对于Transformer的增益随着计算规模的增加而增加，并且在最佳模型尺寸下质量遵循幂律。我们还从经验上验证了Primer可以放入不同的代码库，从而显著加快训练速度，而无需进行额外的调优。例如，在500米的参数大小下，Primer改进了C4自回归语言建模的原始T5架构，将训练成本降低了4倍。此外，降低的训练成本意味着Primer需要更少的计算来达到目标一次性性能。例如，在类似GPT-3 XL的1.9B参数配置中，Primer使用1/3的训练计算来实现与Transformer相同的一次性性能。我们在T5中公开了我们的模型和一些比较，以帮助提高再现性。
摘要：Large Transformer models have been central to recent advances in natural language processing. The training and inference costs of these models, however, have grown rapidly and become prohibitively expensive. Here we aim to reduce the costs of Transformers by searching for a more efficient variant. Compared to previous approaches, our search is performed at a lower level, over the primitives that define a Transformer TensorFlow program. We identify an architecture, named Primer, that has a smaller training cost than the original Transformer and other variants for auto-regressive language modeling. Primer's improvements can be mostly attributed to two simple modifications: squaring ReLU activations and adding a depthwise convolution layer after each Q, K, and V projection in self-attention. Experiments show Primer's gains over Transformer increase as compute scale grows and follow a power law with respect to quality at optimal model sizes. We also verify empirically that Primer can be dropped into different codebases to significantly speed up training without additional tuning. For example, at a 500M parameter size, Primer improves the original T5 architecture on C4 auto-regressive language modeling, reducing the training cost by 4X. Furthermore, the reduced training cost means Primer needs much less compute to reach a target one-shot performance. For instance, in a 1.9B parameter configuration similar to GPT-3 XL, Primer uses 1/3 of the training compute to achieve the same one-shot performance as Transformer. We open source our models and several comparisons in T5 to help with reproducibility.
【2】 Boosting Transformers for Job Expression Extraction and Classification in a Low-Resource Setting
标题：在低资源环境下增强用于作业表达式提取和分类的转换器
链接：https://arxiv.org/abs/2109.08597
机构： Bosch Center for Artificial Intelligence, Robert-Bosch-Campus , Renningen, Germany, Spoken Language Systems (LSV), Saarbr¨ucken Graduate School of Computer Science, Saarland Informatics Campus, Saarland University, Saarbr¨ucken, Germany
备注：Published at IberLEF 2021. Best system of the NER and CLASS tracks of the MEDDOPROF shared task
摘要：在本文中，我们探讨了在低资源环境下Transformer模型的可能改进。特别是，我们提出了解决MEDDOPROF竞赛三个子任务中前两个子任务的方法，即西班牙语临床文本中工作表达的提取和分类。作为既不是语言专家也不是领域专家，我们使用多语言XLM-R transformer模型进行实验，并将这些低资源信息提取任务作为序列标记问题来处理。我们探索领域和语言自适应预训练、迁移学习和策略数据片段，以增强transformer模型。我们的结果表明，与微调的XLM-R模型相比，使用这些方法可以大大提高5.3个F1点。我们的最佳车型在前两项任务中分别达到83.2和79.3 F1。
摘要：In this paper, we explore possible improvements of transformer models in a low-resource setting. In particular, we present our approaches to tackle the first two of three subtasks of the MEDDOPROF competition, i.e., the extraction and classification of job expressions in Spanish clinical texts. As neither language nor domain experts, we experiment with the multilingual XLM-R transformer model and tackle these low-resource information extraction tasks as sequence-labeling problems. We explore domain- and language-adaptive pretraining, transfer learning and strategic datasplits to boost the transformer model. Our results show strong improvements using these methods by up to 5.3 F1 points compared to a fine-tuned XLM-R model. Our best models achieve 83.2 and 79.3 F1 for the first two tasks, respectively.
【3】 From Known to Unknown: Knowledge-guided Transformer for Time-Series Sales Forecasting in Alibaba
标题：从已知到未知：阿里巴巴时间序列销售预测的知识导向Transformer
链接：https://arxiv.org/abs/2109.08381
机构：Alibaba Group
备注：8 pages, 7 figure
摘要：时间序列预测（TSF）在许多实际应用中基本上是必需的，例如电力消耗规划和销售预测。在电子商务中，准确的时间序列销售预测（TSSF）可以显著提高经济效益。TSSF在电子商务领域的目标是预测未来数百万种产品的销售额。产品的趋势和季节性变化很大，促销活动严重影响销售。除了上述困难之外，我们还可以提前知道一些未来的知识，除了历史统计。此类未来知识可能反映未来促销活动对当前销售的影响，并有助于实现更高的准确性。然而，大多数现有的TSF方法仅基于历史信息预测未来。在这项工作中，我们弥补了对未来知识的遗漏。除了引入未来知识进行预测外，我们还提出了基于双向Transformer的智能Transformer，它可以利用历史信息、当前因素和未来知识预测未来的销售额。具体来说，我们设计了一个知识引导的自我注意层，该层使用已知知识的一致性来引导定时信息的传输。提出了面向未来的训练策略，使模型更加注重未来知识的利用。在四个公共基准数据集和天猫一个拟议的大规模工业数据集上进行的大量实验表明，Aliformer的性能远远优于最先进的TSF方法。Aliformer已部署用于天猫行业桌面上的商品选择，数据集将在批准后发布。
摘要：Time series forecasting (TSF) is fundamentally required in many real-world applications, such as electricity consumption planning and sales forecasting. In e-commerce, accurate time-series sales forecasting (TSSF) can significantly increase economic benefits. TSSF in e-commerce aims to predict future sales of millions of products. The trend and seasonality of products vary a lot, and the promotion activity heavily influences sales. Besides the above difficulties, we can know some future knowledge in advance except for the historical statistics. Such future knowledge may reflect the influence of the future promotion activity on current sales and help achieve better accuracy. However, most existing TSF methods only predict the future based on historical information. In this work, we make up for the omissions of future knowledge. Except for introducing future knowledge for prediction, we propose Aliformer based on the bidirectional Transformer, which can utilize the historical information, current factor, and future knowledge to predict future sales. Specifically, we design a knowledge-guided self-attention layer that uses known knowledge's consistency to guide the transmission of timing information. And the future-emphasized training strategy is proposed to make the model focus more on the utilization of future knowledge. Extensive experiments on four public benchmark datasets and one proposed large-scale industrial dataset from Tmall demonstrate that Aliformer can perform much better than state-of-the-art TSF methods. Aliformer has been deployed for goods selection on Tmall Industry Tablework, and the dataset will be released upon approval.
GAN|对抗|攻击|生成相关(1篇)
【1】 Hard to Forget: Poisoning Attacks on Certified Machine Unlearning
标题：难以忘记：对认证机器遗忘的毒化攻击
链接：https://arxiv.org/abs/2109.08266
机构：School of Computing and Information Systems, University of Melbourne, Department of Computer Science, Amherst College
摘要：擦除权要求从组织持有的数据中删除用户信息，严格的解释延伸到下游产品，如学习模型。用省略的特定用户数据从头开始重新训练完全消除了它对结果模型的影响，但计算成本很高。机器“取消学习”降低了完全再训练所产生的成本：相反，模型是增量更新的，可能仅在近似误差累积时才需要再训练。在未学习和再训练模型的不可区分性方面，在隐私保障方面取得了快速进展，但目前的形式主义并未对计算设置实际界限。在本文中，我们将演示攻击者如何利用这一疏忽，重点介绍由机器遗忘引入的一个新的攻击面。我们认为攻击者旨在提高计算成本的数据删除。我们推导并实证研究了一个针对认证机器遗忘的中毒攻击，其中策略性设计的训练数据在移除时触发完全再训练。
摘要：The right to erasure requires removal of a user's information from data held by organizations, with rigorous interpretations extending to downstream products such as learned models. Retraining from scratch with the particular user's data omitted fully removes its influence on the resulting model, but comes with a high computational cost. Machine "unlearning" mitigates the cost incurred by full retraining: instead, models are updated incrementally, possibly only requiring retraining when approximation errors accumulate. Rapid progress has been made towards privacy guarantees on the indistinguishability of unlearned and retrained models, but current formalisms do not place practical bounds on computation. In this paper we demonstrate how an attacker can exploit this oversight, highlighting a novel attack surface introduced by machine unlearning. We consider an attacker aiming to increase the computational cost of data removal. We derive and empirically investigate a poisoning attack on certified machine unlearning where strategically designed training data triggers complete retraining when removed.
半/弱/无/有监督|不确定性|主动学习(5篇)
【1】 Self-Supervised Neural Architecture Search for Imbalanced Datasets
标题：非平衡数据集的自监督神经结构搜索
链接：https://arxiv.org/abs/2109.08580
摘要：神经架构搜索（NAS）在经过精心策划的带有注释标签的数据集上进行训练时，可提供最先进的结果。然而，对来自不同科学领域（如医学领域）的从业者来说，注释数据甚至拥有均衡数量的样本可能是一种奢侈。为此，我们提出了一个基于NAS的框架，该框架有三方面的贡献：（a）我们关注自我监督的场景，即不需要标签来确定体系结构，（b）我们假设数据集是不平衡的，（c）我们设计的每个组件都能够在资源受限的设置上运行，即在单个GPU上运行（例如Google Colab）。我们的组件建立在自我监督学习~\citep{zbontar2021barlow}，自我监督NAS ~\citep{kaplan 2020self}的最新发展之上，并将其扩展到不平衡数据集的情况。我们在一个不平衡版本的CIFAR-10，我们证明了我们提出的方法优于标准神经网络，同时使用的参数少了$27 \倍。为了验证我们在自然不平衡数据集上的假设，我们还对ChestMNIST和COVID-19 X射线进行了实验。结果表明，我们提出的方法可以用于不平衡数据集tasets，但它可以在单个GPU上完全运行。代码可用\href{https://github.com/TimofeevAlex/ssnas_imbalanced}{这里}。
摘要：Neural Architecture Search (NAS) provides state-of-the-art results when trained on well-curated datasets with annotated labels. However, annotating data or even having balanced number of samples can be a luxury for practitioners from different scientific fields, e.g., in the medical domain. To that end, we propose a NAS-based framework that bears the threefold contributions: (a) we focus on the self-supervised scenario, i.e., where no labels are required to determine the architecture, and (b) we assume the datasets are imbalanced, (c) we design each component to be able to run on a resource constrained setup, i.e., on a single GPU (e.g. Google Colab). Our components build on top of recent developments in self-supervised learning~\citep{zbontar2021barlow}, self-supervised NAS~\citep{kaplan2020self} and extend them for the case of imbalanced datasets. We conduct experiments on an (artificially) imbalanced version of CIFAR-10 and we demonstrate our proposed method outperforms standard neural networks, while using $27\times$ less parameters. To validate our assumption on a naturally imbalanced dataset, we also conduct experiments on ChestMNIST and COVID-19 X-ray. The results demonstrate how the proposed method can be used in imbalanced datasets, while it can be fully run on a single GPU. Code is available \href{https://github.com/TimofeevAlex/ssnas_imbalanced}{here}.
【2】 Integrating Deep Reinforcement and Supervised Learning to Expedite Indoor Mapping
标题：深度强化与有监督学习相结合加速室内测图
链接：https://arxiv.org/abs/2109.08490
备注：Submitted to ICRA-22 conference (September 14th, 2021)
摘要：解决了绘制室内环境地图的挑战。用于解决运动规划问题的典型启发式算法是基于边界的方法，在环境完全未知的情况下尤其有效。然而，在环境建筑特征的先验统计数据可用的情况下，这种算法可能远远不是最优的。此外，随着暴露面积的增加，它们的计算时间可能会大幅增加。在本文中，我们提出了两种方法来克服这些缺点。一种是使用深度强化学习来训练运动规划器。第二个是包含一个预先训练的生成性深层神经网络，作为map预测器。每一种方法都有助于通过使用学习到的环境结构统计数据来改进决策，并且这两种方法都作为神经网络实现，确保了恒定的计算时间。我们表明，与基于边界的运动规划相比，将这两种方法结合起来可以缩短映射时间达75%。
摘要：The challenge of mapping indoor environments is addressed. Typical heuristic algorithms for solving the motion planning problem are frontier-based methods, that are especially effective when the environment is completely unknown. However, in cases where prior statistical data on the environment's architectonic features is available, such algorithms can be far from optimal. Furthermore, their calculation time may increase substantially as more areas are exposed. In this paper we propose two means by which to overcome these shortcomings. One is the use of deep reinforcement learning to train the motion planner. The second is the inclusion of a pre-trained generative deep neural network, acting as a map predictor. Each one helps to improve the decision making through use of the learned structural statistics of the environment, and both, being realized as neural networks, ensure a constant calculation time. We show that combining the two methods can shorten the mapping time, compared to frontier-based motion planning, by up to 75%.
【3】 Improving Regression Uncertainty Estimation Under Statistical Change
标题：统计变化下回归不确定度估计的改进
链接：https://arxiv.org/abs/2109.08213
机构：Massachusetts Institute of Technology,Raytheon BBN Technologies
摘要：虽然深度神经网络在广泛的现实世界问题中表现出色并取得成功，但估计其预测不确定性仍然是一项具有挑战性的任务。为了应对这一挑战，我们提出并实现了一个基于贝叶斯验证度量（BVM）框架的回归不确定性估计损失函数，同时使用集成学习。对分布数据的一系列实验表明，该方法与现有的先进方法相比具有一定的竞争力。此外，对非分布数据的实验表明，该方法对统计变化具有较强的鲁棒性，具有较好的预测能力。
摘要：While deep neural networks are highly performant and successful in a wide range of real-world problems, estimating their predictive uncertainty remains a challenging task. To address this challenge, we propose and implement a loss function for regression uncertainty estimation based on the Bayesian Validation Metric (BVM) framework while using ensemble learning. A series of experiments on in-distribution data show that the proposed method is competitive with existing state-of-the-art methods. In addition, experiments on out-of-distribution data show that the proposed method is robust to statistical change and exhibits superior predictive capability.
【4】 Adaptive Hierarchical Dual Consistency for Semi-Supervised Left Atrium Segmentation on Cross-Domain Data
标题：跨域数据半监督左心房分割的自适应分层对偶一致性
链接：https://arxiv.org/abs/2109.08311
机构：Yang, Senior Member, IEEE
摘要：半监督学习在标记数据不足的左心房分割模型学习中具有重要意义。将半监督学习推广到跨领域数据对于进一步提高模型的鲁棒性具有重要意义。然而，不同数据域之间广泛存在的分布差异和样本不匹配阻碍了半监督学习的推广。在这项研究中，我们提出了一种自适应分层双重一致性（AHDC）算法，用于跨域数据的半监督LA分割，从而缓解了这些问题。AHDC主要由双向对抗性推理模块（BAI）和分层双一致性学习模块（HDC）组成。BAI克服了分布的差异和两个不同域之间的样本不匹配。它主要通过对两个映射网络进行对抗学习，通过互适配获得两个匹配域。HDC研究了一种基于获得的匹配域的跨域半监督分割的分层双学习范式。它主要构建两个双重建模网络，用于挖掘域内和域间的互补信息。对于域内学习，将一致性约束应用于双建模目标，以利用互补的建模信息。对于域间学习，将一致性约束应用于由两个双重建模网络建模的LAs，以利用不同数据域之间的互补知识。我们在来自不同中心的四个3D晚期钆增强心脏MR（LGE-CMR）数据集和一个3D CT数据集上展示了我们提出的AHDC的性能。与其他最先进的方法相比，我们提出的AHDC实现了更高的分割精度，这表明了它在跨域半监督LA分割中的能力。
摘要：Semi-supervised learning provides great significance in left atrium (LA) segmentation model learning with insufficient labelled data. Generalising semi-supervised learning to cross-domain data is of high importance to further improve model robustness. However, the widely existing distribution difference and sample mismatch between different data domains hinder the generalisation of semi-supervised learning. In this study, we alleviate these problems by proposing an Adaptive Hierarchical Dual Consistency (AHDC) for the semi-supervised LA segmentation on cross-domain data. The AHDC mainly consists of a Bidirectional Adversarial Inference module (BAI) and a Hierarchical Dual Consistency learning module (HDC). The BAI overcomes the difference of distributions and the sample mismatch between two different domains. It mainly learns two mapping networks adversarially to obtain two matched domains through mutual adaptation. The HDC investigates a hierarchical dual learning paradigm for cross-domain semi-supervised segmentation based on the obtained matched domains. It mainly builds two dual-modelling networks for mining the complementary information in both intra-domain and inter-domain. For the intra-domain learning, a consistency constraint is applied to the dual-modelling targets to exploit the complementary modelling information. For the inter-domain learning, a consistency constraint is applied to the LAs modelled by two dual-modelling networks to exploit the complementary knowledge among different data domains. We demonstrated the performance of our proposed AHDC on four 3D late gadolinium enhancement cardiac MR (LGE-CMR) datasets from different centres and a 3D CT dataset. Compared to other state-of-the-art methods, our proposed AHDC achieved higher segmentation accuracy, which indicated its capability in the cross-domain semi-supervised LA segmentation.
【5】 Assessments of model-form uncertainty using Gaussian stochastic weight averaging for fluid-flow regression
标题：基于高斯随机加权平均的流体回归模型形式不确定度评定
链接：https://arxiv.org/abs/2109.08248
机构：Department of Mechanical Engineering, Keio University, Yokohama,-, Japan, Department of Mechanical and Aerospace Engineering, University of California, Los Angeles, CA, Mathematics and Computer Science Division, Argonne National Laboratory, Illinois , USA
摘要：我们使用高斯随机加权平均（SWAG）来评估与流体流动相关的基于神经网络的函数近似相关的模型形式不确定性。给定训练数据和恒定的学习率，SWAG近似每个权重的后验高斯分布。有了这个分布，它就能够创建多个具有不同采样权重组合的模型，这些组合可用于获得集合预测。此类集合的平均值可被视为“平均估计”，而其标准偏差可用于构造“置信区间”，这使我们能够对神经网络的训练过程进行不确定性量化（UQ）。我们利用具有代表性的基于神经网络的函数逼近任务处理以下情况：（i）二维圆柱尾迹；（ii）DayMET数据集（北美每日最高温度）；（iii）三维方形圆柱尾迹；和（iv）城市流量，以评估当前想法对广泛复杂数据集的普遍性。无论网络结构如何，基于SWAG的UQ都可以应用，因此，我们证明了该方法对两类神经网络的适用性：（i）结合卷积神经网络（CNN）和多层感知器（MLP）从稀疏传感器重建全局场；以及（ii）利用二维CNN从截面数据进行远场状态估计。我们发现，SWAG可以从模型形式不确定性的角度获得物理上可解释的置信区间估计。这种能力支持它用于科学和工程领域的广泛问题。
摘要：We use Gaussian stochastic weight averaging (SWAG) to assess the model-form uncertainty associated with neural-network-based function approximation relevant to fluid flows. SWAG approximates a posterior Gaussian distribution of each weight, given training data, and a constant learning rate. Having access to this distribution, it is able to create multiple models with various combinations of sampled weights, which can be used to obtain ensemble predictions. The average of such an ensemble can be regarded as the `mean estimation', whereas its standard deviation can be used to construct `confidence intervals', which enable us to perform uncertainty quantification (UQ) with regard to the training process of neural networks. We utilize representative neural-network-based function approximation tasks for the following cases: (i) a two-dimensional circular-cylinder wake; (ii) the DayMET dataset (maximum daily temperature in North America); (iii) a three-dimensional square-cylinder wake; and (iv) urban flow, to assess the generalizability of the present idea for a wide range of complex datasets. SWAG-based UQ can be applied regardless of the network architecture, and therefore, we demonstrate the applicability of the method for two types of neural networks: (i) global field reconstruction from sparse sensors by combining convolutional neural network (CNN) and multi-layer perceptron (MLP); and (ii) far-field state estimation from sectional data with two-dimensional CNN. We find that SWAG can obtain physically-interpretable confidence-interval estimates from the perspective of model-form uncertainty. This capability supports its use for a wide range of problems in science and engineering.
迁移|Zero/Few/One-Shot|自适应(1篇)
【1】 AdaLoss: A computationally-efficient and provably convergent adaptive gradient method
标题：AdaLoss：一种计算高效且可证明收敛的自适应梯度法
链接：https://arxiv.org/abs/2109.08282
机构：University of Chicago, University of Texas at Austin, University of Washington
备注：arXiv admin note: text overlap with arXiv:1902.07111
摘要：我们提出了一种计算友好的自适应学习速率调度“AdaLoss”，它直接利用损失函数的信息来调整梯度下降法中的步长。我们证明了该调度在线性回归中具有线性收敛性。此外，在两层参数化神经网络的背景下，我们在非凸区域上提供了线性收敛保证。如果两层网络中第一个隐藏层的宽度足够大（多项式），则AdaLoss在多项式时间内稳健地收敛到全局最小值。通过考虑文本澄清和控制问题的策略梯度在LSTM模型中的应用，我们对理论结果进行了数值验证，并扩展了数值实验的范围。
摘要：We propose a computationally-friendly adaptive learning rate schedule, "AdaLoss", which directly uses the information of the loss function to adjust the stepsize in gradient descent methods. We prove that this schedule enjoys linear convergence in linear regression. Moreover, we provide a linear convergence guarantee over the non-convex regime, in the context of two-layer over-parameterized neural networks. If the width of the first-hidden layer in the two-layer networks is sufficiently large (polynomially), then AdaLoss converges robustly \emph{to the global minimum} in polynomial time. We numerically verify the theoretical results and extend the scope of the numerical experiments by considering applications in LSTM models for text clarification and policy gradients for control problems.
强化学习(5篇)
【1】 Decentralized Global Connectivity Maintenance for Multi-Robot Navigation: A Reinforcement Learning Approach
标题：多机器人导航的分散全局连通性维护：一种强化学习方法
链接：https://arxiv.org/abs/2109.08536
机构： [ 1] have proposed a potential-basedmethod towards circle obstacles avoidance with connectivityThe authors are with the School of Computer Science and Engineeringand the School of Eletronics and Communication Engineering, Sun Yat-senUniversity
备注：7 pages, 15 figures
摘要：多机器人导航的连通性维护问题在多机器人应用中具有挑战性。这项工作研究如何在未知环境中导航多机器人团队，同时保持连接。我们提出了一种强化学习（RL）方法来开发分散策略，该策略在多个机器人之间共享。给定距离传感器测量值和其他机器人的位置，该策略旨在生成导航控制命令，并保持机器人团队的全球连通性。我们将连接性问题作为约束纳入RL框架，并引入行为克隆以降低策略优化的探索复杂性。该策略通过随机模拟场景中多个机器人收集的所有过渡数据进行优化。我们通过比较连接约束和行为克隆的不同组合来验证所提出方法的有效性。我们还表明，我们的策略可以推广到模拟和完整机器人实验中看不见的场景。
摘要：The problem of multi-robot navigation of connectivity maintenance is challenging in multi-robot applications. This work investigates how to navigate a multi-robot team in unknown environments while maintaining connectivity. We propose a reinforcement learning (RL) approach to develop a decentralized policy, which is shared among multiple robots. Given range sensor measurements and the positions of other robots, the policy aims to generate control commands for navigation and preserve the global connectivity of the robot team. We incorporate connectivity concerns into the RL framework as constraints and introduce behavior cloning to reduce the exploration complexity of policy optimization. The policy is optimized with all transition data collected by multiple robots in random simulated scenarios. We validate the effectiveness of the proposed approach by comparing different combinations of connectivity constraints and behavior cloning. We also show that our policy can generalize to unseen scenarios in both simulation and holonomic robots experiments.
【2】 Coordinated Random Access for Industrial IoT With Correlated Traffic By Reinforcement-Learning
标题：基于强化学习的业务量相关工业物联网协同随机接入
链接：https://arxiv.org/abs/2109.08389
机构：Department of Information Engineering, University of Padova, Italy.
备注：6 pages, conference paper (GLOBECOM2021)
摘要：我们提出了一种用于工业物联网（IIoT）场景的协调随机接入方案，其中机器类型设备（MTD）生成零星的相关流量。例如，当外部事件在多个MTD同时触发数据生成时，就会发生这种情况。时间被划分为帧，每个帧被划分为时隙，每个MTD随机选择一个时隙用于（重新）传输，概率密度函数（pdf）特定于MTD和当前重新传输的次数。PDF是局部优化的，以最小化数据包冲突的概率。该优化问题被建模为一个不完全信息的重复马尔可夫博弈，在每个MTD上使用线性奖励不作为算法，该算法可证明收敛到确定性（次优）时隙分配。我们将我们的解决方案与时隙ALOHA和最小-最大成对相关随机接入方案进行了比较，结果表明我们的方法在中等流量强度的情况下实现了更高的网络吞吐量。
摘要：We propose a coordinated random access scheme for industrial internet-of-things (IIoT) scenarios, with machine-type devices (MTDs) generating sporadic correlated traffic. This occurs, e.g., when external events trigger data generation at multiple MTDs simultaneously. Time is divided into frames, each split into slots and each MTD randomly selects one slot for (re)transmission, with probability density functions (PDFs) specific of both the MTD and the number of the current retransmission. PDFs are locally optimized to minimize the probability of packet collision. The optimization problem is modeled as a repeated Markov game with incomplete information, and the linear reward-inaction algorithm is used at each MTD, which provably converges to a deterministic (suboptimal) slot assignment. We compare our solution with both the slotted ALOHA and the min-max pairwise correlation random access schemes, showing that our approach achieves a higher network throughput with moderate traffic intensity.
【3】 Accelerating Offline Reinforcement Learning Application in Real-Time Bidding and Recommendation: Potential Use of Simulation
标题：加速离线强化学习在实时投标和推荐中的应用：仿真的潜在用途
链接：https://arxiv.org/abs/2109.08331
机构： Japan and Tokyo Institute of Technology, Japan and Cornell University
备注：SimuRec workshop at RecSys2021
摘要：在在线广告推荐系统（RecSys）和实时竞价（RTB）中，我们经常尝试使用bandit和强化学习（RL）技术优化顺序决策。在这些应用程序中，离线强化学习（离线RL）和离线策略评估（OPE）是有益的，因为它们只使用记录的数据实现安全策略优化，而无需任何风险在线交互。在这篇论文中，我们探讨了使用仿真加速离线RL和OPE的实际研究的潜力，特别是在RecSys和RTB中。具体而言，我们将讨论模拟如何帮助我们进行离线RL和OPE的实证研究。我们主张在离线RL和OPE的实证研究中应有效地使用模拟。为了反驳只使用真实世界数据的实验更可取的反诉，我们首先指出了真实世界实验中潜在的风险和再现性问题。然后，我们描述了如何通过模拟来解决这些问题。此外，我们还展示了如何结合真实世界和基于模拟的实验的优点来捍卫我们的立场。最后，我们还提出了一个开放的挑战，以进一步促进RecSys和RTB中离线RL和OPE在公共仿真平台方面的实际研究。作为这个问题的一个可能的解决方案，我们展示了我们正在进行的开源项目及其潜在的用例。我们相信，为离线RL和OPE构建和利用基于仿真的评估平台将对RecSys和RTB社区产生极大的兴趣和相关性。
摘要：In recommender systems (RecSys) and real-time bidding (RTB) for online advertisements, we often try to optimize sequential decision making using bandit and reinforcement learning (RL) techniques. In these applications, offline reinforcement learning (offline RL) and off-policy evaluation (OPE) are beneficial because they enable safe policy optimization using only logged data without any risky online interaction. In this position paper, we explore the potential of using simulation to accelerate practical research of offline RL and OPE, particularly in RecSys and RTB. Specifically, we discuss how simulation can help us conduct empirical research of offline RL and OPE. We take a position to argue that we should effectively use simulations in the empirical research of offline RL and OPE. To refute the counterclaim that experiments using only real-world data are preferable, we first point out the underlying risks and reproducibility issue in real-world experiments. Then, we describe how these issues can be addressed by using simulations. Moreover, we show how to incorporate the benefits of both real-world and simulation-based experiments to defend our position. Finally, we also present an open challenge to further facilitate practical research of offline RL and OPE in RecSys and RTB, with respect to public simulation platforms. As a possible solution for the issue, we show our ongoing open source project and its potential use case. We believe that building and utilizing simulation-based evaluation platforms for offline RL and OPE will be of great interest and relevance for the RecSys and RTB community.
【4】 Reinforcement Learning on Encrypted Data
标题：加密数据的强化学习
链接：https://arxiv.org/abs/2109.08236
机构：University of Bologna, University College London, The Alan Turing Institute
摘要：由于数据本身的敏感性，强化学习（RL）在现实领域的应用日益增多，导致了隐私保护技术的发展。大多数现有的工作都集中在差异隐私上，在差异隐私中，信息以透明的方式透露给代理，代理的学习模型应该能够抵抗向恶意第三方泄漏的信息。由于仅使用加密数据可能被共享的用例，例如来自敏感站点的信息，在这项工作中，我们考虑输入本身是敏感的并且不能被揭示的场景。我们开发了MDP框架的一个简单扩展，它提供了状态加密。我们提出了一个初步的，实验研究如何在离散和连续状态空间的环境中执行DQN代理训练加密状态。我们的结果表明，即使存在非确定性加密，代理仍然能够在小的状态空间中学习，但在更复杂的环境中性能会崩溃。
摘要：The growing number of applications of Reinforcement Learning (RL) in real-world domains has led to the development of privacy-preserving techniques due to the inherently sensitive nature of data. Most existing works focus on differential privacy, in which information is revealed in the clear to an agent whose learned model should be robust against information leakage to malicious third parties. Motivated by use cases in which only encrypted data might be shared, such as information from sensitive sites, in this work we consider scenarios in which the inputs themselves are sensitive and cannot be revealed. We develop a simple extension to the MDP framework which provides for the encryption of states. We present a preliminary, experimental study of how a DQN agent trained on encrypted states performs in environments with discrete and continuous state spaces. Our results highlight that the agent is still capable of learning in small state spaces even in presence of non-deterministic encryption, but performance collapses in more complex environments.
【5】 RAPID-RL: A Reconfigurable Architecture with Preemptive-Exits for Efficient Deep-Reinforcement Learning
标题：RAPID-RL：一种高效深度强化学习的抢占出口可重构体系结构
链接：https://arxiv.org/abs/2109.08231
摘要：当今的深度强化学习（RL）系统在构建超越人类水平的智能代理方面显示出巨大的潜力。然而，与底层深层神经网络（DNN）相关的计算复杂性导致了耗电的实现。这使得深度RL系统不适合部署在资源受限的边缘设备上。为了应对这一挑战，我们提出了一种具有抢占式出口的可重构体系结构，用于高效的深度RL（RAPID-RL）。RAPID-RL可根据输入的难度水平有条件激活DNN层。这允许在推理过程中动态调整计算工作量，同时保持竞争性能。我们通过增加深度Q网络（DQN）和能够生成中间预测以及相关置信度得分的分支来实现这一点。我们还提出了一种新的训练方法，用于在动态RL环境中学习动作和分支可信度分数。我们的实验在开源无人机模拟器（PEDRA）上评估了Atari 2600游戏任务和真实无人机导航任务的拟议框架。我们表明，与没有任何分支的基线DQN相比，RAPID-RL在Atari（无人机导航）任务上的性能保持在0.88倍（0.91倍）以上的同时，产生了0.34倍（0.25倍）的操作数（OPS）。OPS的减少导致了快速高效的推理，这对于资源受限的边缘非常有利，因为在这些边缘，必须以最少的计算量做出快速决策。
摘要：Present-day Deep Reinforcement Learning (RL) systems show great promise towards building intelligent agents surpassing human-level performance. However, the computational complexity associated with the underlying deep neural networks (DNNs) leads to power-hungry implementations. This makes deep RL systems unsuitable for deployment on resource-constrained edge devices. To address this challenge, we propose a reconfigurable architecture with preemptive exits for efficient deep RL (RAPID-RL). RAPID-RL enables conditional activation of DNN layers based on the difficulty level of inputs. This allows to dynamically adjust the compute effort during inference while maintaining competitive performance. We achieve this by augmenting a deep Q-network (DQN) with side-branches capable of generating intermediate predictions along with an associated confidence score. We also propose a novel training methodology for learning the actions and branch confidence scores in a dynamic RL setting. Our experiments evaluate the proposed framework for Atari 2600 gaming tasks and a realistic Drone navigation task on an open-source drone simulator (PEDRA). We show that RAPID-RL incurs 0.34x (0.25x) number of operations (OPS) while maintaining performance above 0.88x (0.91x) on Atari (Drone navigation) tasks, compared to a baseline-DQN without any side-branches. The reduction in OPS leads to fast and efficient inference, proving to be highly beneficial for the resource-constrained edge where making quick decisions with minimal compute is essential.
元学习(1篇)
【1】 Automatic prior selection for meta Bayesian optimization with a case study on tuning deep neural network optimizers
标题：元贝叶斯优化的自动先验选择--以深度神经网络优化器的调整为例
链接：https://arxiv.org/abs/2109.08215
摘要：深度神经网络的性能对各种元参数的选择非常敏感，例如优化器参数和模型超参数。然而，调整好这些参数通常需要大量且昂贵的实验。贝叶斯优化（BO）是一种有效解决此类昂贵的超参数优化问题的原则性方法。BO性能的关键是指定和细化函数上的分布，该分布用于推断正在优化的底层函数的最优值。在这项工作中，我们考虑的情况下，我们有数据从类似的功能，使我们能够指定更严格的分布先验。具体地说，我们专注于为训练神经网络调整优化器参数这一常见但可能代价高昂的任务。基于Wang et al.（2018）提出的meta-BO方法，我们开发了一些实际改进，这些改进（a）通过利用多个任务上的调整结果来提高其性能，而无需在所有任务中对相同的元参数点进行观察，（b）对于我们方法的特殊情况，保留其遗憾边界。因此，我们为连续优化器参数的迭代优化提供了一致的BO解。为了在真实的模型训练设置中验证我们的方法，我们通过在流行的图像和文本数据集以及蛋白质序列数据集上训练几万个接近最先进的模型配置，收集了一个大型多任务超参数调整数据集。我们的结果表明，平均而言，我们的方法能够定位好的超参数，其效率至少是最佳竞争方法的3倍。
摘要 ：The performance of deep neural networks can be highly sensitive to the choice of a variety of meta-parameters, such as optimizer parameters and model hyperparameters. Tuning these well, however, often requires extensive and costly experimentation. Bayesian optimization (BO) is a principled approach to solve such expensive hyperparameter tuning problems efficiently. Key to the performance of BO is specifying and refining a distribution over functions, which is used to reason about the optima of the underlying function being optimized. In this work, we consider the scenario where we have data from similar functions that allows us to specify a tighter distribution a priori. Specifically, we focus on the common but potentially costly task of tuning optimizer parameters for training neural networks. Building on the meta BO method from Wang et al. (2018), we develop practical improvements that (a) boost its performance by leveraging tuning results on multiple tasks without requiring observations for the same meta-parameter points across all tasks, and (b) retain its regret bound for a special case of our method. As a result, we provide a coherent BO solution for iterative optimization of continuous optimizer parameters. To verify our approach in realistic model training setups, we collected a large multi-task hyperparameter tuning dataset by training tens of thousands of configurations of near-state-of-the-art models on popular image and text datasets, as well as a protein sequence dataset. Our results show that on average, our method is able to locate good hyperparameters at least 3 times more efficiently than the best competing methods.
符号|符号学习(2篇)
【1】 Conversational Multi-Hop Reasoning with Neural Commonsense Knowledge and Symbolic Logic Rules
标题：基于神经常识和符号逻辑规则的对话式多跳推理
链接：https://arxiv.org/abs/2109.08544
机构：Facebook, EPFL, University of Washington, Carnegie Mellon University
备注：Appearing in the 2021 Conference on Empirical Methods in Natural Language Processing (EMNLP)
摘要：会话代理所面临的一个挑战是，它们无法识别用户命令的未声明假设，由于人类的常识，这项任务对人类来说微不足道。在本文中，我们提出了一个会话代理Zero-Shot常识推理系统，试图实现这一点。我们的推理机从满足if-（state），then-（action），because-（goal）通用模板的用户命令中发现了未声明的假设。我们的推理机使用最先进的基于Transformer的生成性常识知识库（KB）作为推理的背景知识源。我们提出了一种新的迭代知识查询机制，从神经知识库中提取多跳推理链，该机制使用符号逻辑规则来显著减少搜索空间。与迄今为止收集的任何知识库类似，我们的常识知识库容易丢失知识。因此，我们建议使用新的动态问题生成策略，通过对话从人类用户那里获取缺失的知识，该策略生成并向人类用户呈现情境化的查询。我们通过对人类用户的用户研究对模型进行评估，与SOTA相比，成功率高出35%。
摘要：One of the challenges faced by conversational agents is their inability to identify unstated presumptions of their users' commands, a task trivial for humans due to their common sense. In this paper, we propose a zero-shot commonsense reasoning system for conversational agents in an attempt to achieve this. Our reasoner uncovers unstated presumptions from user commands satisfying a general template of if-(state), then-(action), because-(goal). Our reasoner uses a state-of-the-art transformer-based generative commonsense knowledge base (KB) as its source of background knowledge for reasoning. We propose a novel and iterative knowledge query mechanism to extract multi-hop reasoning chains from the neural KB which uses symbolic logic rules to significantly reduce the search space. Similar to any KBs gathered to date, our commonsense KB is prone to missing knowledge. Therefore, we propose to conversationally elicit the missing knowledge from human users with our novel dynamic question generation strategy, which generates and presents contextualized queries to human users. We evaluate the model with a user study with human users that achieves a 35% higher success rate compared to SOTA.
【2】 Natlog: a Lightweight Logic Programming Language with a Neuro-symbolic Touch
标题：Natlog：一种具有神经符号触觉的轻量级逻辑程序设计语言
链接：https://arxiv.org/abs/2109.08291
机构：This work is licensed under the, Creative Commons Attribution License., Natlog: a Lightweight Logic Programming Language with a, Neuro-symbolic Touch, University of North Texas, Texas, USA
备注：None
摘要：我们介绍了一种轻量级逻辑编程语言Natlog，它共享Prolog的统一驱动执行模型，但具有简化的语法和语义。我们的概念验证Natlog实现紧密嵌入在基于Python的深度学习生态系统中，重点关注基础术语数据集的内容驱动索引。作为我们符号索引算法的替代，同样的功能可以委托给神经网络，为Natlog的解析引擎提供基本事实。我们的开源实现可以作为Python包在https://pypi.org/project/natlog/ .
摘要：We introduce Natlog, a lightweight Logic Programming language, sharing Prolog's unification-driven execution model, but with a simplified syntax and semantics. Our proof-of-concept Natlog implementation is tightly embedded in the Python-based deep-learning ecosystem with focus on content-driven indexing of ground term datasets. As an overriding of our symbolic indexing algorithm, the same function can be delegated to a neural network, serving ground facts to Natlog's resolution engine. Our open-source implementation is available as a Python package at https://pypi.org/project/natlog/ .
医学相关(4篇)
【1】 Slot Filling for Biomedical Information Extraction
标题：用于生物医学信息提取的缝隙填充
链接：https://arxiv.org/abs/2109.08564
机构：Healx, Cambridge, UK
摘要：信息提取（IE）“从文本”是指从非结构化文本中提取结构化知识的任务。该任务通常由一系列子任务组成，如命名实体识别和关系提取。获取实体和关系类型特定的训练数据是上述子任务中的一个主要瓶颈。在这项工作中，我们提出了一种时隙填充方法针对生物医学IE的任务，有效地取代了对实体和关系特定训练数据的需求，允许处理Zero-Shot设置。我们遵循最近提出的范例，将基于转换器的bi编码器、密集通道检索与基于转换器的读卡器模型耦合，以从生物医学文本中提取关系。我们mble是一个用于检索和阅读理解的生物医学插槽填充数据集，并进行了一系列实验，证明我们的方法优于许多简单的基线。我们还评估了我们的方法在标准和零炮设置下的端到端。我们的工作为如何解决生物医学IE t提供了一个新的视角在缺乏相关训练数据的情况下询问。我们的代码、模型和预训练数据可在https://github.com/healx/biomed-slot-filling.
摘要：Information Extraction (IE) from text refers to the task of extracting structured knowledge from unstructured text. The task typically consists of a series of sub-tasks such as Named Entity Recognition and Relation Extraction. Sourcing entity and relation type specific training data is a major bottleneck in the above sub-tasks.In this work we present a slot filling approach to the task of biomedical IE, effectively replacing the need for entity and relation-specific training data, allowing to deal with zero-shot settings. We follow the recently proposed paradigm of coupling a Tranformer-based bi-encoder, Dense Passage Retrieval, with a Transformer-based reader model to extract relations from biomedical text. We assemble a biomedical slot filling dataset for both retrieval and reading comprehension and conduct a series of experiments demonstrating that our approach outperforms a number of simpler baselines. We also evaluate our approach end-to-end for standard as well as zero-shot settings. Our work provides a fresh perspective on how to solve biomedical IE tasks, in the absence of relevant training data. Our code, models and pretrained data are available at https://github.com/healx/biomed-slot-filling.
【2】 An Interpretable Framework for Drug-Target Interaction with Gated Cross Attention
标题：具有门控交叉注意的药物-靶点相互作用的可解释框架
链接：https://arxiv.org/abs/2109.08360
机构：Deargen Inc., Seoul, South Korea
备注：16 pages, Proceedings of Machine Learning for Healthcare, 2021 (MLHC'21)
摘要：电子版药物靶点相互作用预测（DTI）对于药物发现具有重要意义，因为它可以大大缩短药物开发过程的时间和成本。具体而言，基于深度学习的DTI方法在预测的准确性和低成本方面已显示出有希望的结果。然而，他们很少关注预测结果的可解释性以及药物和靶点之间的特征级相互作用。在这项研究中，我们提出了一个新的解释框架，可以为相互作用位点提供合理的线索。为此，我们精心设计了一个门控交叉注意机制，通过构建这些特征之间的显式相互作用，交叉关注药物和目标特征。该方法中的门控功能使神经模型能够关注整个药物和蛋白质序列上的显著区域，而该功能的副产品，即注意力图，可以作为可解释的因素。在两个DTI数据集中的实验结果表明了该方法的有效性。此外，我们还表明，门控交叉注意可以对突变作出敏感反应，这一结果可以为识别针对突变蛋白的新药提供见解。
摘要：In silico prediction of drug-target interactions (DTI) is significant for drug discovery because it can largely reduce timelines and costs in the drug development process. Specifically, deep learning-based DTI approaches have been shown promising results in terms of accuracy and low cost for the prediction. However, they pay little attention to the interpretability of their prediction results and feature-level interactions between a drug and a target. In this study, we propose a novel interpretable framework that can provide reasonable cues for the interaction sites. To this end, we elaborately design a gated cross-attention mechanism that crossly attends drug and target features by constructing explicit interactions between these features. The gating function in the method enables neural models to focus on salient regions over entire sequences of drugs and proteins, and the byproduct from the function, which is the attention map, could serve as interpretable factors. The experimental results show the efficacy of the proposed method in two DTI datasets. Additionally, we show that gated cross-attention can sensitively react to the mutation, and this result could provide insights into the identification of novel drugs targeting mutant proteins.
【3】 A review of deep learning methods for MRI reconstruction
标题：MRI重建中的深度学习方法综述
链接
：https://arxiv.org/abs/2109.08618
摘要：随着深度学习在广泛应用中的成功，基于神经网络的机器学习技术在加速磁共振成像（MRI）采集和重建策略方面受到了广泛关注。受计算机视觉和图像处理的深度学习技术启发，许多想法已成功地应用于非线性图像重建，其核心是加速MRI的压缩传感。鉴于该领域的快速发展，有必要巩固和总结文献中报道的大量深度学习方法，以便更好地了解该领域的总体情况。本文概述了基于神经网络的方法的最新发展，这些方法专门用于改进并行成像。从基于k空间的重建方法的经典观点出发，给出了并行MRI的一般背景和介绍。介绍了引入改进正则化器的基于图像域的技术以及基于k空间的方法，这些方法侧重于使用神经网络的更好的插值策略。虽然该领域正在迅速发展，每年发表数千篇论文，但在本综述中，我们试图涵盖在公开数据集上表现良好的方法的广泛类别。还讨论了局限性和开放性问题，并审查了最近为社区创建开放数据集和基准所做的努力。
摘要：Following the success of deep learning in a wide range of applications, neural network-based machine-learning techniques have received significant interest for accelerating magnetic resonance imaging (MRI) acquisition and reconstruction strategies. A number of ideas inspired by deep learning techniques for computer vision and image processing have been successfully applied to nonlinear image reconstruction in the spirit of compressed sensing for accelerated MRI. Given the rapidly growing nature of the field, it is imperative to consolidate and summarize the large number of deep learning methods that have been reported in the literature, to obtain a better understanding of the field in general. This article provides an overview of the recent developments in neural-network based approaches that have been proposed specifically for improving parallel imaging. A general background and introduction to parallel MRI is also given from a classical view of k-space based reconstruction methods. Image domain based techniques that introduce improved regularizers are covered along with k-space based methods which focus on better interpolation strategies using neural networks. While the field is rapidly evolving with thousands of papers published each year, in this review, we attempt to cover broad categories of methods that have shown good performance on publicly available data sets. Limitations and open problems are also discussed and recent efforts for producing open data sets and benchmarks for the community are examined.
【4】 Stereo Video Reconstruction Without Explicit Depth Maps for Endoscopic Surgery
标题：无显式深度图的内窥镜手术立体视频重建
链接：https://arxiv.org/abs/2109.08227
机构：Center for Data Science, New York University, Department of Neurosurgery, Langone Health, New York University
备注：9 pages, 5 figures
摘要：我们介绍了用于微创手术视频的立体视频重建或等效的2D-to-3D视频转换的任务。我们通过改变输入（单帧与多个连续帧）、损耗函数（MSE、MAE或感知损耗）和网络架构，为此任务设计并实现了一系列基于端到端U-Net的解决方案。我们通过调查10位专家来评估这些解决方案，这些专家是例行进行内窥镜手术的外科医生。我们进行了两项独立的读者研究：一项评估单个帧，另一项评估在VR耳机上播放的完全重建的3D视频。在第一个读者研究中，U-Net的一种变体，它将多个连续视频帧作为输入，并输出缺失的视图，性能最佳。我们从这一结果中得出两个结论。首先，来自多个过去帧的运动信息对于重建立体视觉至关重要。第二，所提出的U-Net变体确实可以利用这些运动信息来解决这项任务。第二项研究的结果进一步证实了所提出的U-Net变体的有效性。外科医生报告说，他们可以成功地从重建的3D视频剪辑中感知深度。他们还明确表示，与原始2D视频相比，他们更喜欢重建的3D视频。这两项读者研究有力地支持了针对微创手术视频的立体重建任务的实用性，并表明深度学习是完成这项任务的一种有希望的方法。最后，我们确定了两个自动度量，LPIP和DIST，它们与专家判断密切相关，可以作为后者在未来研究中的代理。
摘要：We introduce the task of stereo video reconstruction or, equivalently, 2D-to-3D video conversion for minimally invasive surgical video. We design and implement a series of end-to-end U-Net-based solutions for this task by varying the input (single frame vs. multiple consecutive frames), loss function (MSE, MAE, or perceptual losses), and network architecture. We evaluate these solutions by surveying ten experts - surgeons who routinely perform endoscopic surgery. We run two separate reader studies: one evaluating individual frames and the other evaluating fully reconstructed 3D video played on a VR headset. In the first reader study, a variant of the U-Net that takes as input multiple consecutive video frames and outputs the missing view performs best. We draw two conclusions from this outcome. First, motion information coming from multiple past frames is crucial in recreating stereo vision. Second, the proposed U-Net variant can indeed exploit such motion information for solving this task. The result from the second study further confirms the effectiveness of the proposed U-Net variant. The surgeons reported that they could successfully perceive depth from the reconstructed 3D video clips. They also expressed a clear preference for the reconstructed 3D video over the original 2D video. These two reader studies strongly support the usefulness of the proposed task of stereo reconstruction for minimally invasive surgical video and indicate that deep learning is a promising approach to this task. Finally, we identify two automatic metrics, LPIPS and DISTS, that are strongly correlated with expert judgement and that could serve as proxies for the latter in future studies.
蒸馏|知识提取(1篇)
【1】 Distilling Linguistic Context for Language Model Compression
标题：用于语言模型压缩的语境提取
链接：https://arxiv.org/abs/2109.08359
机构：KAIST, Daejeon, South Korea, AITRICS, Seoul, South Korea
备注：EMNLP 2021. Code: this https URL
摘要：最近语言表征学习的成功背后是一个计算量大且记忆密集的神经网络。知识提取是在资源匮乏的环境中部署如此庞大的语言模型的一项主要技术，它可以不受限制地将所学的单个单词表示的知识转移。在本文中，受最近观察到的语言表征相对定位且整体上具有更多语义知识的启发，我们提出了一个新的语言表征学习知识提取目标，该目标通过两种类型的跨表征关系传递上下文知识：单词关系和层转换关系。与语言模型的其他最新蒸馏技术不同，我们的上下文蒸馏对教师和学生之间的架构更改没有任何限制。我们验证了我们的方法在具有挑战性的语言理解任务基准测试中的有效性，不仅在各种规模的体系结构中，而且结合最近提出的自适应规模修剪方法DynaBERT。
摘要：A computationally expensive and memory intensive neural network lies behind the recent success of language representation learning. Knowledge distillation, a major technique for deploying such a vast language model in resource-scarce environments, transfers the knowledge on individual word representations learned without restrictions. In this paper, inspired by the recent observations that language representations are relatively positioned and have more semantic knowledge as a whole, we present a new knowledge distillation objective for language representation learning that transfers the contextual knowledge via two types of relationships across representations: Word Relation and Layer Transforming Relation. Unlike other recent distillation techniques for the language models, our contextual distillation does not have any restrictions on architectural changes between teacher and student. We validate the effectiveness of our method on challenging benchmarks of language understanding tasks, not only in architectures of various sizes, but also in combination with DynaBERT, the recently proposed adaptive size pruning method.
推荐(2篇)
【1】 Context-aware Retail Product Recommendation with Regularized Gradient Boosting
标题：基于正则化梯度提升的情境感知零售产品推荐
链接：https://arxiv.org/abs/2109.08561
机构：Razorthink Inc,USA
备注：Accepted to the FARFETCH Fashion Recommendations Challenge Workshop, ECML-PKDD 2021
摘要：在FARFETCH时尚推荐挑战赛中，参与者需要预测在推荐印象中向用户展示各种产品的顺序。数据分两个阶段提供——验证阶段和测试阶段。验证阶段有一个带标签的训练集，其中包含一个二进制列，指示er一个产品是否被点击。数据集包括500多万个推荐事件、45万个产品和23万个独特用户。它代表FARFETCH平台实际用户的真实、无偏见但匿名的互动。最终评估是根据第二阶段的表现进行的。共有167名参与者参加了挑战，我们在最终评估中获得了第6名，测试集的MRR为0.4658。我们设计了一个独特的上下文感知系统，该系统考虑了产品与用户上下文的相似性，以便更有效地对产品进行排名。评估后，我们能够使用测试集的MRR为0.4784，这将使我们处于第三位。
摘要 ：In the FARFETCH Fashion Recommendation challenge, the participants needed to predict the order in which various products would be shown to a user in a recommendation impression. The data was provided in two phases - a validation phase and a test phase. The validation phase had a labelled training set that contained a binary column indicating whether a product has been clicked or not. The dataset comprises over 5,000,000 recommendation events, 450,000 products and 230,000 unique users. It represents real, unbiased, but anonymised, interactions of actual users of the FARFETCH platform. The final evaluation was done according to the performance in the second phase. A total of 167 participants participated in the challenge, and we secured the 6th rank during the final evaluation with an MRR of 0.4658 on the test set. We have designed a unique context-aware system that takes the similarity of a product to the user context into account to rank products more effectively. Post evaluation, we have been able to fine-tune our approach with an MRR of 0.4784 on the test set, which would have placed us at the 3rd position.
【2】 Multi-Level Visual Similarity Based Personalized Tourist Attraction Recommendation Using Geo-Tagged Photos
标题：基于多级视觉相似度的地标照片个性化旅游景点推荐
链接：https://arxiv.org/abs/2109.08275
机构：College of Computer Science and Technology, Zhejiang University, Hangzhou , China
备注：17 pages, 4 figures
摘要：基于地理标记照片的旅游景点推荐可以从用户拍摄的照片中发现用户的旅游偏好，从而向用户推荐合适的旅游景点。然而，现有的基于视觉内容的方法不能充分利用照片的用户和景点信息来提取视觉特征，也不能区分不同照片的意义。在本文中，我们提出了基于多层次视觉相似性的个性化旅游景点推荐方法。MEIN利用照片的视觉内容和交互行为数据来获得用户和旅游景点的最终嵌入，然后用于预测访问概率。具体来说，通过交叉照片的用户和景点信息，定义了四个视觉相似性级别，并引入了相应的五重损失来嵌入照片的视觉内容。此外，为了捕捉不同照片的意义，我们利用自我注意机制获取用户和旅游景点的视觉表征。我们在一个从Flickr抓取的数据集上进行了实验，实验结果证明了该方法的优越性。
摘要：Geo-tagged photo based tourist attraction recommendation can discover users' travel preferences from their taken photos, so as to recommend suitable tourist attractions to them. However, existing visual content based methods cannot fully exploit the user and tourist attraction information of photos to extract visual features, and do not differentiate the significances of different photos. In this paper, we propose multi-level visual similarity based personalized tourist attraction recommendation using geo-tagged photos (MEAL). MEAL utilizes the visual contents of photos and interaction behavior data to obtain the final embeddings of users and tourist attractions, which are then used to predict the visit probabilities. Specifically, by crossing the user and tourist attraction information of photos, we define four visual similarity levels and introduce a corresponding quintuplet loss to embed the visual contents of photos. In addition, to capture the significances of different photos, we exploit the self-attention mechanism to obtain the visual representations of users and tourist attractions. We conducted experiments on a dataset crawled from Flickr, and the experimental results proved the advantage of this method.
聚类(1篇)
【1】 Discriminative Similarity for Data Clustering
标题：用于数据聚类的判别相似度
链接：https://arxiv.org/abs/2109.08675
机构：School of Computing and Augmented Intelligence, Arizona State University, S Mill RD, Tempe, AZ , USA, Cognitive Computing Lab, Baidu Research, NE ,th St. Bellevue, WA , USA
备注：arXiv admin note: substantial text overlap with arXiv:1709.01231
摘要：基于相似性的聚类方法根据数据之间的两两相似性将数据分为若干类，而两两相似性对聚类的性能至关重要。在本文中，我们提出了一种利用判别相似性（CDS）进行聚类的新方法。CDS从每个数据分区学习一个无监督的基于相似性的分类器，并通过最小化与数据分区相关联的学习分类器的泛化误差来搜索数据的最佳分区。通过Rademacher复杂度的泛化分析，将基于无监督相似性的分类器的泛化误差界表示为不同类别数据之间的判别相似性之和。证明了核密度分类的积分平方误差界也可以导出判别相似性。为了评估所提出的判别相似性的性能，我们提出了一种新的聚类方法，使用一个核作为相似性函数，CDS通过无监督核分类（CDSK），实验结果证明了其有效性。
摘要：Similarity-based clustering methods separate data into clusters according to the pairwise similarity between the data, and the pairwise similarity is crucial for their performance. In this paper, we propose Clustering by Discriminative Similarity (CDS), a novel method which learns discriminative similarity for data clustering. CDS learns an unsupervised similarity-based classifier from each data partition, and searches for the optimal partition of the data by minimizing the generalization error of the learnt classifiers associated with the data partitions. By generalization analysis via Rademacher complexity, the generalization error bound for the unsupervised similarity-based classifier is expressed as the sum of discriminative similarity between the data from different classes. It is proved that the derived discriminative similarity can also be induced by the integrated squared error bound for kernel density classification. In order to evaluate the performance of the proposed discriminative similarity, we propose a new clustering method using a kernel as the similarity function, CDS via unsupervised kernel classification (CDSK), with its effectiveness demonstrated by experimental results.
联邦学习|隐私保护|加密(3篇)
【1】 Enforcing fairness in private federated learning via the modified method of differential multipliers
标题：利用改进的差分乘子法实现私有联合学习中的公平性
链接：https://arxiv.org/abs/2109.08604
机构：KTH Royal Institute of Technology – Information Science and Engineering (ISE)
摘要：具有差异隐私的联合学习（或称私有联合学习）提供了一种在尊重用户隐私的同时训练机器学习模型的策略。然而，差异隐私会不成比例地降低模型在代表性不足的群体上的性能，因为分布的这些部分在存在噪声的情况下很难学习。现有的在机器学习模型中实现公平性的方法都考虑了集中式设置，其中算法可以访问用户的数据。本文介绍了一种在私有联合学习中实现组公平性的算法，在这种算法中，用户的数据不会离开他们的设备。首先，本文将改进的差分乘数法推广到具有公平约束的经验风险最小化，从而提供了一种在中心环境中强制执行公平性的算法。然后，将该算法扩展到私有联邦学习环境。提出的算法FPFL在成人数据集的联邦版本和FEMNIST数据集的“不公平”版本上进行了测试。在这些数据集上的实验表明，私有联合学习如何在训练模型中加剧不公平，以及FPFL如何能够缓解这种不公平。
摘要：Federated learning with differential privacy, or private federated learning, provides a strategy to train machine learning models while respecting users' privacy. However, differential privacy can disproportionately degrade the performance of the models on under-represented groups, as these parts of the distribution are difficult to learn in the presence of noise. Existing approaches for enforcing fairness in machine learning models have considered the centralized setting, in which the algorithm has access to the users' data. This paper introduces an algorithm to enforce group fairness in private federated learning, where users' data does not leave their devices. First, the paper extends the modified method of differential multipliers to empirical risk minimization with fairness constraints, thus providing an algorithm to enforce fairness in the central setting. Then, this algorithm is extended to the private federated learning setting. The proposed algorithm, FPFL, is tested on a federated version of the Adult dataset and an "unfair" version of the FEMNIST dataset. The experiments on these datasets show how private federated learning accentuates unfairness in the trained models, and how FPFL is able to mitigate such unfairness.
【2】 Comfetch: Federated Learning of Large Networks on Memory-Constrained Clients via Sketching
标题：ComFetch：基于草图的大型网络在内存受限客户端上的联合学习
链接：https://arxiv.org/abs/2109.08346
机构：Huang , University of Maryland, College Park
摘要：联邦学习的一个流行应用是使用许多客户端来训练深度神经网络，其参数保存在中央服务器上。虽然最近的努力集中于降低通信复杂性，但现有算法假设每个参与客户端都能够下载当前和完整的参数集，这可能不是一个实际的假设，这取决于客户端（如移动设备）的内存限制。在这项工作中，我们提出了一种新的算法Comfetch，它允许客户端通过计数草图使用全局架构的压缩版本来训练大型网络，从而降低通信和本地内存成本。我们提供了一个理论上的收敛保证，并通过实验证明，通过在其草图对应物上训练联邦代理，可以学习大型网络，例如深卷积网络和LSTM。与最先进的FetchSGD和经典的FedAvg相比，生成的全局模型显示出具有竞争力的测试精度，这两种模型都要求客户端下载完整的体系结构。
摘要 ：A popular application of federated learning is using many clients to train a deep neural network, the parameters of which are maintained on a central server. While recent efforts have focused on reducing communication complexity, existing algorithms assume that each participating client is able to download the current and full set of parameters, which may not be a practical assumption depending on the memory constraints of clients such as mobile devices. In this work, we propose a novel algorithm Comfetch, which allows clients to train large networks using compressed versions of the global architecture via Count Sketch, thereby reducing communication and local memory costs. We provide a theoretical convergence guarantee and experimentally demonstrate that it is possible to learn large networks, such as a deep convolutional network and an LSTM, through federated agents training on their sketched counterparts. The resulting global models exhibit competitive test accuracy when compared against the state-of-the-art FetchSGD and the classical FedAvg, both of which require clients to download the full architecture.
【3】 Achieving Model Fairness in Vertical Federated Learning
标题：垂直联合学习中模型公平性的实现
链接：https://arxiv.org/abs/2109.08344
机构： Pei is with the School of Computing Science
备注：15 pages, 2 figures
摘要：垂直联合学习（VFL）使具有非重叠特征的多个企业能够在不公开其私有数据和模型参数的情况下增强其机器学习模型，最近受到了越来越多的关注。与其他机器学习算法类似，VFL存在公平性问题，即学习模型可能对具有敏感属性的组具有不公平的歧视性。为了解决这个问题，我们在这项工作中提出了一个公平的VFL框架。首先，我们系统地阐述了VFL中的训练公平模型问题，其中学习任务被建模为一个约束优化问题。为了解决它的联合方式，我们考虑它的等价对偶形式，并开发了异步梯度坐标下降上升算法，其中每个数据方执行每个通信轮的多个并行本地更新，以有效地减少通信回合数。我们证明了在温和的条件下，该算法在$\mathcal{O}（\delta^{-4}）$通信轮中找到对偶目标的$\delta$平稳点。最后，在三个基准数据集上的大量实验证明了我们的方法在训练公平模型方面的优越性能。
摘要：Vertical federated learning (VFL), which enables multiple enterprises possessing non-overlapped features to strengthen their machine learning models without disclosing their private data and model parameters, has received increasing attention lately. Similar to other machine learning algorithms, VFL suffers from fairness issues, i.e., the learned model may be unfairly discriminatory over the group with sensitive attributes. To tackle this problem, we propose a fair VFL framework in this work. First, we systematically formulate the problem of training fair models in VFL, where the learning task is modeled as a constrained optimization problem. To solve it in a federated manner, we consider its equivalent dual form and develop an asynchronous gradient coordinate-descent ascent algorithm, where each data party performs multiple parallelized local updates per communication round to effectively reduce the number of communication rounds. We prove that the algorithm finds a $\delta$-stationary point of the dual objective in $\mathcal{O}(\delta^{-4})$ communication rounds under mild conditions. Finally, extensive experiments on three benchmark datasets demonstrate the superior performance of our method in training fair models.
推理|分析|理解|解释(3篇)
【1】 A Fairness Analysis on Private Aggregation of Teacher Ensembles
标题：教师群体私募的公平性分析
链接：https://arxiv.org/abs/2109.08630
机构：Syracuse University
摘要：教师集合的私有聚合（PATE）是一个重要的私有机器学习框架。它将作为教师使用的多个学习模型与学生模型相结合，学生模型通过在教师中进行嘈杂投票来学习预测输出。所得到的模型满足差异隐私，并且在半监督环境下或当希望保护数据标签时，在学习高质量的私有模型方面已被证明是有效的。本文询问这种隐私保护框架是否会引入或加剧偏见和不公平性，并表明PATE会引入个人和个人群体之间的准确性差异。本文分析了哪些算法和数据属性导致了不相称的影响，为什么这些方面会对不同群体产生不相称的影响，并提出了减轻这些影响的指导方针。在几个数据集和设置上对所提出的方法进行了评估。
摘要：The Private Aggregation of Teacher Ensembles (PATE) is an important private machine learning framework. It combines multiple learning models used as teachers for a student model that learns to predict an output chosen by noisy voting among the teachers. The resulting model satisfies differential privacy and has been shown effective in learning high-quality private models in semisupervised settings or when one wishes to protect the data labels. This paper asks whether this privacy-preserving framework introduces or exacerbates bias and unfairness and shows that PATE can introduce accuracy disparity among individuals and groups of individuals. The paper analyzes which algorithmic and data properties are responsible for the disproportionate impacts, why these aspects are affecting different groups disproportionately, and proposes guidelines to mitigate these effects. The proposed approach is evaluated on several datasets and settings.
【2】 Micro-architectural Analysis of a Learned Index
标题：一种学习型索引的微体系结构分析
链接：https://arxiv.org/abs/2109.08495
机构：IT University of Copenhagen, Denmark
备注：Under submission
摘要：自2018年《学习索引结构案例》出版以来，针对不同领域和不同功能的学习索引的研究有所增加。虽然一些研究已经证明了学习索引作为传统索引结构（如B+树）的替代方法的有效性，但以前的工作倾向于关注更高级别的性能指标，如吞吐量和索引大小。在本文中，我们的目标是更深入地挖掘和研究与传统索引相比，学习索引在微观体系结构层面上的表现。更具体地说，我们关注先前提出的学习索引结构ALEX，它是一种基于树的内存索引结构，由机器学习模型的层次结构组成。与学习索引的原始方案不同，ALEX从头开始设计，允许更新和插入。因此，它可以使用学习的索引实现更动态的工作负载。在这项工作中，我们对ALEX进行了微观架构分析，并将其行为与基于树的索引结构（不基于学习模型，即ART和B+树）进行比较。我们的结果表明，ALEX受到内存暂停的限制，主要是由于最后一级缓存中的数据丢失而导致的暂停。与ART和B+树相比，ALEX在不同的工作负载中表现出更少的暂停和更低的每个指令值周期。另一方面，在ALEX中处理绑定外插入所需的指令量可以显著增加每个请求所需的指令量（10倍），以适应写操作繁重的工作负载。然而，微体系结构行为表明，指令占用空间的增加表现出较高的指令级并行性，因此不会对总体执行时间产生负面影响。
摘要：Since the publication of The Case for Learned Index Structures in 2018, there has been a rise in research that focuses on learned indexes for different domains and with different functionalities. While the effectiveness of learned indexes as an alternative to traditional index structures such as B+Trees have already been demonstrated by several studies, previous work tend to focus on higher-level performance metrics such as throughput and index size. In this paper, our goal is to dig deeper and investigate how learned indexes behave at a micro-architectural level compared to traditional indexes. More specifically, we focus on previously proposed learned index structure ALEX, which is a tree-based in-memory index structure that consists of a hierarchy of machine learned models. Unlike the original proposal for learned indexes, ALEX is designed from the ground up to allow updates and inserts. Therefore, it enables more dynamic workloads using learned indexes. In this work, we perform a micro-architectural analysis of ALEX and compare its behavior to the tree-based index structures that are not based on learned models, i.e., ART and B+Tree. Our results show that ALEX is bound by memory stalls, mainly stalls due to data misses from the last-level cache. Compared to ART and B+Tree, ALEX exhibits fewer stalls and a lower cycles-per-instruction value across different workloads. On the other hand, the amount of instructions required to handle out-of-bound inserts in ALEX can increase the instructions needed per request significantly (10X) for write-heavy workloads. However, the micro-architectural behavior shows that this increase in the instruction footprint exhibit high instruction-level parallelism, and, therefore, does not negatively impact the overall execution time.
【3】 TS-MULE: Local Interpretable Model-Agnostic Explanations for Time Series Forecast Models
标题：TS-MULE：时间序列预测模型的局部可解释模型不可知性解释
链接：https://arxiv.org/abs/2109.08438
机构：University of Konstanz, Germany
备注：8 pages, 2 pages references, Advances in Interpretable Machine Learning and Artificial Intelligence (AIMLAI) at ECML/PKDD 2021
摘要：时间序列预测是一项艰巨的任务，从天气预测到故障预测，黑箱模型实现了最先进的性能。但是，不能保证理解和调试。我们提出TS-MULE，这是一种局部替代模型解释方法，专门用于扩展LIME方法的时间序列。我们的扩展LIME使用各种方法分割和扰动时间序列数据。在我们的扩展中，我们提出了六种时间序列抽样分割方法，以提高代理属性的质量，并在三种深度学习模型结构和三种常见的多元时间序列数据集上展示了它们的性能。
摘要：Time series forecasting is a demanding task ranging from weather to failure forecasting with black-box models achieving state-of-the-art performances. However, understanding and debugging are not guaranteed. We propose TS-MULE, a local surrogate model explanation method specialized for time series extending the LIME approach. Our extended LIME works with various ways to segment and perturb the time series data. In our extension, we present six sampling segmentation approaches for time series to improve the quality of surrogate attributions and demonstrate their performances on three deep learning model architectures and three common multivariate time series datasets.
检测相关(2篇)
【1】 An open GPS trajectory dataset and benchmark for travel mode detection
标题：一种开放的GPS轨迹数据集和出行模式检测基准
链接：https://arxiv.org/abs/2109.08527
机构：Center for Spatial Information Science, The University of Tokyo,-,-, Kashiwanoha, Kashiwa, Chiba ,-, Japan, SUSTech-UTokyo Joint Research Center on Super Smart City, Department of Computer
摘要：行程模式检测一直是GPS轨迹相关处理领域的研究热点。以前的学者已经发展了许多数学方法来提高检测的准确性。在这些研究中，几乎所有的方法都需要地面真实数据集进行训练。大量的研究选择通过定制的方式采集GPS轨迹数据集进行训练。目前，没有标记有行驶模式的开放式GPS数据集。如果存在一个模型，它不仅可以节省模型开发的大量工作，还可以帮助比较模型的性能。在本研究中，我们提出并开放带有行程模式和基准的GPS轨迹数据集，用于行程模式检测。该数据集由日本的7名独立志愿者收集，涵盖一个完整月的时间段。旅行方式从步行到铁路。一部分例行程序在不同的时间段重复行驶，以体验不同的道路和行驶条件。我们还提供了一个案例研究，以区分大规模GPS轨迹数据集中的步行和自行车出行。
摘要：Travel mode detection has been a hot topic in the field of GPS trajectory-related processing. Former scholars have developed many mathematical methods to improve the accuracy of detection. Among these studies, almost all of the methods require ground truth dataset for training. A large amount of the studies choose to collect the GPS trajectory dataset for training by their customized ways. Currently, there is no open GPS dataset marked with travel mode. If there exists one, it will not only save a lot of efforts in model developing, but also help compare the performance of models. In this study, we propose and open GPS trajectory dataset marked with travel mode and benchmark for the travel mode detection. The dataset is collected by 7 independent volunteers in Japan and covers the time period of a complete month. The travel mode ranges from walking to railway. A part of routines are traveled repeatedly in different time slots to experience different road and travel conditions. We also provide a case study to distinguish the walking and bike trips in a massive GPS trajectory dataset.
【2】 Towards agricultural autonomy: crop row detection under varying field conditions using deep learning
标题：走向农业自主化：利用深度学习在不同田间条件下进行作物行检测
链接：https://arxiv.org/abs/2109.08247
机构：LincolnInstituteforAgri-FoodTechnology, UniversityofLincoln
备注：This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
摘要：本文提出了一种新的度量方法，用于评估基于深度学习的语义分割方法在田间机器人遇到的不同田间条件下对作物行检测的鲁棒性。在不同的现场条件下，使用了一个包含十个主要类别的数据集进行测试。比较了这些条件对作物行检测角度精度的影响。采用深卷积编解码网络，利用RGB输入图像预测裁剪行掩模。然后将预测的遮罩发送到后处理算法以提取裁剪行。研究发现，深度学习模型对阴影和作物生长阶段具有鲁棒性，而在阳光直射下，该模型的性能有所下降，当使用新的度量进行评估时，杂草密度增加，作物行中的条纹和不连续性增加。
摘要：This paper presents a novel metric to evaluate the robustness of deep learning based semantic segmentation approaches for crop row detection under different field conditions encountered by a field robot. A dataset with ten main categories encountered under various field conditions was used for testing. The effect on these conditions on the angular accuracy of crop row detection was compared. A deep convolutional encoder decoder network is implemented to predict crop row masks using RGB input images. The predicted mask is then sent to a post processing algorithm to extract the crop rows. The deep learning model was found to be robust against shadows and growth stages of the crop while the performance was reduced under direct sunlight, increasing weed density, tramlines and discontinuities in crop rows when evaluated with the novel metric.
分类|识别(2篇)
【1】 Beyond Average Performance -- exploring regions of deviating performance for black box classification models
标题：超越平均性能--探索黑盒分类模型的偏离性能区域
链接：https://arxiv.org/abs/2109.08216
机构：Received: date Accepted: date
摘要：机器学习模型在不同类型的环境中越来越流行。这主要是因为他们能够实现人类专家在新的大数据时代难以比拟的预测性能水平。随着使用量的增长，对模型预测的责任和理解的要求也随之增加。然而，最成功的模型（如集成、深度学习）的复杂程度正成为这一努力的一大障碍，因为这些模型本质上是黑匣子。在本文中，我们描述了两种通用的方法，它们可以用来解释任何黑盒分类模型的预期性能。这些方法具有很高的实用性，因为它们提供了一种方法，可以以可解释的方式揭示和描述模型预期性能与其平均行为显著偏离的情况。这对于那些成本高昂的决策是由模型预测驱动的应用程序来说可能至关重要，因为它可以用来警告最终用户在某些特定情况下不要使用模型。
摘要：Machine learning models are becoming increasingly popular in different types of settings. This is mainly caused by their ability to achieve a level of predictive performance that is hard to match by human experts in this new era of big data. With this usage growth comes an increase of the requirements for accountability and understanding of the models' predictions. However, the degree of sophistication of the most successful models (e.g. ensembles, deep learning) is becoming a large obstacle to this endeavour as these models are essentially black boxes. In this paper we describe two general approaches that can be used to provide interpretable descriptions of the expected performance of any black box classification model. These approaches are of high practical relevance as they provide means to uncover and describe in an interpretable way situations where the models are expected to have a performance that deviates significantly from their average behaviour. This may be of critical relevance for applications where costly decisions are driven by the predictions of the models, as it can be used to warn end users against the usage of the models in some specific cases.
【2】 Policy Choice and Best Arm Identification: Comments on "Adaptive Treatment Assignment in Experiments for Policy Choice"
链接：https://arxiv.org/abs/2109.08229
机构：AI Lab, CyberAgent, Inc., Stern School of Business, New York University, Fox School of Business, Temple University
摘要：本文的目的是将Kasy和Sautmann（2021）提出的“政策选择”问题与机器学习中的强盗文学前沿联系起来。我们讨论如何将政策选择问题框架化，使其与所谓的“最佳武器识别”（BAI）问题相同。通过连接文献，我们发现Kasy和Sautmann（2021）中讨论的政策选择算法的渐近最优性是文献中一个长期存在的开放问题。不幸的是，这种联系突出了主定理的几个主要问题。特别是，我们证明了Kasy和Sautmann（2021）中的定理1是错误的。我们发现，定理1的陈述（1）和（2）的证明是不正确的，尽管陈述本身可能是正确的，尽管要纠正这些陈述并不容易。另一方面，陈述（3）及其证明是错误的，我们利用强盗文献中现有的理论结果证明了这一点。由于这个问题非常重要，在过去十年中，在土匪社区中引起了极大的兴趣，我们对白族文学的最新发展进行了回顾。我们希望这有助于突出与经济问题的相关性，并刺激计量经济学界的方法论和理论发展。
摘要：The purpose of this paper is to connect the "policy choice" problem, proposed in Kasy and Sautmann (2021), to the frontiers of the bandit literature in machine learning. We discuss how the policy choice problem can be framed in a way such that it is identical to what is called the "best arm identification" (BAI) problem. By connecting the literature, we identify that the asymptotic optimality of policy choice algorithms tackled in Kasy and Sautmann (2021) is a long-standing open question in the literature. Unfortunately, this connection highlights several major issues with the main theorem. In particular, we show that Theorem 1 in Kasy and Sautmann (2021) is false. We find that the proofs of statements (1) and (2) of Theorem 1 are incorrect, though the statements themselves may be true, though non-trivial to fix. Statement (3), and its proof, on the other hand, is false, which we show by utilizing existing theoretical results in the bandit literature. As this question is critically important, garnering much interest in the last decade within the bandit community, we provide a review of recent developments in the BAI literature. We hope this serves to highlight the relevance to economic problems and stimulate methodological and theoretical developments in the econometric community.
优化|敛散性(3篇)
【1】 Scheduling in Parallel Finite Buffer Systems: Optimal Decisions under Delayed Feedback
标题：并行有限缓冲系统中的调度：延迟反馈下的最优决策
链接：https://arxiv.org/abs/2109.08548
摘要 ：并行排队系统中的调度决策是一个基本问题，是许多计算和通信系统（如数据中心集群中的作业路由、多路径通信和大数据系统）的规模和操作的基础。本质上，调度器将每个到达的作业映射到一个可能的异构服务器，同时以负载平衡、低平均延迟或低丢失率等优化目标为目标。在这里寻找最优调度决策的一个主要困难是调度器仅部分观察其决策的影响，例如，通过服务作业的延迟确认。在本文中，我们提供了一个部分可观测（PO）模型，该模型在有限的延迟确认信息下捕获并行排队系统中的调度决策。我们提出了一个该PO系统的仿真模型，以使用可扩展的蒙特卡罗树搜索算法实时找到近似最优的调度策略。我们的数值结果表明，该策略优于其他有限信息调度策略，如加入最多观测值的变体，并且与全信息策略具有相当的性能，如：加入最短队列、加入最短队列（d）和最短预期延迟。最后，我们展示了我们的方法如何通过使用Kaggle提供的网络数据来优化实时并行处理。
摘要：Scheduling decisions in parallel queuing systems arise as a fundamental problem, underlying the dimensioning and operation of many computing and communication systems, such as job routing in data center clusters, multipath communication, and Big Data systems. In essence, the scheduler maps each arriving job to one of the possibly heterogeneous servers while aiming at an optimization goal such as load balancing, low average delay or low loss rate. One main difficulty in finding optimal scheduling decisions here is that the scheduler only partially observes the impact of its decisions, e.g., through the delayed acknowledgements of the served jobs. In this paper, we provide a partially observable (PO) model that captures the scheduling decisions in parallel queuing systems under limited information of delayed acknowledgements. We present a simulation model for this PO system to find a near-optimal scheduling policy in real-time using a scalable Monte Carlo tree search algorithm. We numerically show that the resulting policy outperforms other limited information scheduling strategies such as variants of Join-the-Most-Observations and has comparable performance to full information strategies like: Join-the-Shortest-Queue, Join-the- Shortest-Queue(d) and Shortest-Expected-Delay. Finally, we show how our approach can optimise the real-time parallel processing by using network data provided by Kaggle.
【2】 CompilerGym: Robust, Performant Compiler Optimization Environments for AI Research
标题：CompilerGym：面向人工智能研究的健壮、高性能编译器优化环境
链接：https://arxiv.org/abs/2109.08267
机构：Facebook
备注：12 pages. Source code available at this https URL
摘要：将人工智能（AI）技术应用于编译器优化的兴趣正在迅速增加，但编译器研究的门槛很高。与其他领域不同的是，编译器和人工智能研究人员无法访问能够快速迭代和开发想法的数据集和框架，而入门需要大量的工程投资。我们需要的是一个简单的、可重用的实验基础设施，用于现实世界的编译器优化任务，它可以作为比较技术的通用基准，并作为加速该领域进展的平台。我们将介绍CompilerGym，一组用于实际编译器优化任务的环境，以及一个向编译器研究人员公开新优化任务的工具包。CompilerGym使任何人都可以通过一个易于使用的包来试验生产编译器优化问题，而不必考虑他们使用编译器的经验。我们基于流行的OpenAI Gym接口，使研究人员能够使用Python和熟悉的API与编译器交互。我们描述了CompilerGym的体系结构和实现，描述了三个包含的编译器环境的优化空间和计算效率，并提供了广泛的经验评估。与以前的工作相比，CompilerGym提供了更大的数据集和优化空间，计算效率提高了27倍，具有容错能力，并且能够检测底层编译器中的再现性错误。为了让任何人都能轻松地使用编译器进行实验——不管他们的背景如何——我们的目标是加速AI和编译器研究领域的进展。
摘要：Interest in applying Artificial Intelligence (AI) techniques to compiler optimizations is increasing rapidly, but compiler research has a high entry barrier. Unlike in other domains, compiler and AI researchers do not have access to the datasets and frameworks that enable fast iteration and development of ideas, and getting started requires a significant engineering investment. What is needed is an easy, reusable experimental infrastructure for real world compiler optimization tasks that can serve as a common benchmark for comparing techniques, and as a platform to accelerate progress in the field. We introduce CompilerGym, a set of environments for real world compiler optimization tasks, and a toolkit for exposing new optimization tasks to compiler researchers. CompilerGym enables anyone to experiment on production compiler optimization problems through an easy-to-use package, regardless of their experience with compilers. We build upon the popular OpenAI Gym interface enabling researchers to interact with compilers using Python and a familiar API. We describe the CompilerGym architecture and implementation, characterize the optimization spaces and computational efficiencies of three included compiler environments, and provide extensive empirical evaluations. Compared to prior works, CompilerGym offers larger datasets and optimization spaces, is 27x more computationally efficient, is fault-tolerant, and capable of detecting reproducibility bugs in the underlying compilers. In making it easy for anyone to experiment with compilers - irrespective of their background - we aim to accelerate progress in the AI and compiler research domains.
【3】 Knowledge is reward: Learning optimal exploration by predictive reward cashing
标题：知识即报酬：通过预测性报酬兑现学习最优探索
链接：https://arxiv.org/abs/2109.08518
机构：Donders Institute, Radboud University, Nijmegen, Netherlands
摘要：智力的一般概念与收集和使用信息的能力之间有着密切的联系。贝叶斯自适应探索理论为训练机器执行复杂的信息收集任务提供了一个有吸引力的优化框架。然而，由此产生的最优控制问题的计算复杂性限制了该理论在主流深度人工智能研究中的推广。在本文中，我们利用贝叶斯自适应问题固有的数学结构，通过使奖励结构更加密集，同时解耦开发和探索策略的学习，从而大大简化问题。这种简化的关键来自交叉价值的新概念（即，在一个环境中，同时根据另一个环境采取最佳行动的价值），我们用它来量化当前可用信息的价值。这导致了一种新的更密集的奖励结构，即“兑现”所有可以从当前信息状态预测的未来奖励。在一组实验中，我们表明，在标准RL算法失败的情况下，该方法可以在不使用成形和启发式奖金的情况下学习具有挑战性的信息收集任务。
摘要：There is a strong link between the general concept of intelligence and the ability to collect and use information. The theory of Bayes-adaptive exploration offers an attractive optimality framework for training machines to perform complex information gathering tasks. However, the computational complexity of the resulting optimal control problem has limited the diffusion of the theory to mainstream deep AI research. In this paper we exploit the inherent mathematical structure of Bayes-adaptive problems in order to dramatically simplify the problem by making the reward structure denser while simultaneously decoupling the learning of exploitation and exploration policies. The key to this simplification comes from the novel concept of cross-value (i.e. the value of being in an environment while acting optimally according to another), which we use to quantify the value of currently available information. This results in a new denser reward structure that "cashes in" all future rewards that can be predicted from the current information state. In a set of experiments we show that the approach makes it possible to learn challenging information gathering tasks without the use of shaping and heuristic bonuses in situations where the standard RL algorithms fail.
其他神经网络|深度学习|模型|建模(9篇)
【1】 Autonomous Vision-based UAV Landing with Collision Avoidance using Deep Learning
标题：基于深度学习的自主视觉无人机着陆避碰
链接：https://arxiv.org/abs/2109.08628
机构：Department of Electrical Engineering and Computer Science, York University, North York, Canada, Department of Earth and Space Science and Engineering
摘要：当多架无人机在同一平台上同时着陆而没有通信时，存在碰撞风险。本文实现了基于视觉的自主着陆，并采用基于深度学习的方法实现了着陆过程中的避碰。
摘要：There is a risk of collision when multiple UAVs land simultaneously without communication on the same platform. This work accomplishes vision-based autonomous landing and uses a deep-learning-based method to realize collision avoidance during the landing process.
【2】 Dynamics-Aware Quality-Diversity for Efficient Learning of Skill Repertoires
标题：有效学习技能套路的动态感知质量多样性
链接：https://arxiv.org/abs/2109.08522
摘要：质量多样性（Quality Diversity，QD）算法是一种功能强大的探索算法，允许机器人发现大量多样且高性能的技能。然而，QD算法样本效率低，需要数百万次评估。在本文中，我们提出了动态感知质量多样性（DA-QD），这是一个通过使用动态模型来提高QD算法样本效率的框架。我们还展示了DA-QD如何用于不断获得新技能。为此，我们根据使用QD进行技能发现时获得的经验，逐步训练深度动力学模型。然后，我们可以在想象中用想象的技能进行量子点探索。我们在三个机器人实验中评估了我们的方法。首先，我们的实验表明，DA-QD在技能发现方面的样本效率是现有QD方法的20倍。第二，我们展示了如何在想象中学习一种全新的技能，从而实现Zero-Shot学习。最后，我们展示了DA-QD在解决长视距导航任务和在现实世界中进行损伤适应方面是如何有用和有效的。视频和源代码可从以下网址获得：https://sites.google.com/view/da-qd.
摘要 ：Quality-Diversity (QD) algorithms are powerful exploration algorithms that allow robots to discover large repertoires of diverse and high-performing skills. However, QD algorithms are sample inefficient and require millions of evaluations. In this paper, we propose Dynamics-Aware Quality-Diversity (DA-QD), a framework to improve the sample efficiency of QD algorithms through the use of dynamics models. We also show how DA-QD can then be used for continual acquisition of new skill repertoires. To do so, we incrementally train a deep dynamics model from experience obtained when performing skill discovery using QD. We can then perform QD exploration in imagination with an imagined skill repertoire. We evaluate our approach on three robotic experiments. First, our experiments show DA-QD is 20 times more sample efficient than existing QD approaches for skill discovery. Second, we demonstrate learning an entirely new skill repertoire in imagination to perform zero-shot learning. Finally, we show how DA-QD is useful and effective for solving a long horizon navigation task and for damage adaptation in the real world. Videos and source code are available at: https://sites.google.com/view/da-qd.
【3】 Online Learning of Network Bottlenecks via Minimax Paths
标题：基于极小极大路径的网络瓶颈在线学习
链接：https://arxiv.org/abs/2109.08467
机构： Chalmers University of Technology, Volvo Car Corporation
摘要：本文研究了通过提取极大极小路径来识别网络中的瓶颈问题。许多现实世界的网络都具有随机权重，而这些权重的全部知识事先并不可用。因此，我们将此任务建模为一个组合半bandit问题，我们应用了一种组合形式的汤普森抽样，并在相应的贝叶斯遗憾上建立了一个上界。由于问题的计算难度，我们设计了一个近似于原始目标的替代问题公式。最后，我们在真实的有向和无向网络上用近似公式对Thompson采样的性能进行了实验评估。
摘要：In this paper, we study bottleneck identification in networks via extracting minimax paths. Many real-world networks have stochastic weights for which full knowledge is not available in advance. Therefore, we model this task as a combinatorial semi-bandit problem to which we apply a combinatorial version of Thompson Sampling and establish an upper bound on the corresponding Bayesian regret. Due to the computational intractability of the problem, we then devise an alternative problem formulation which approximates the original objective. Finally, we experimentally evaluate the performance of Thompson Sampling with the approximate formulation on real-world directed and undirected networks.
【4】 New Students on Sesame Street: What Order-Aware Matrix Embeddings Can Learn from BERT
标题：芝麻街的新学生：顺序感知矩阵嵌入可以从BERT那里学到什么
链接：https://arxiv.org/abs/2109.08449
机构：Kiel University ZBW, Germany, uni-kiel.de, University of Ulm, Germany
备注：10 pages + 6 pages supplementary material
摘要：大规模预训练语言模型（PreLMs）正在所有基准测试中彻底改变自然语言处理。然而，在低资源或大规模应用程序中，它们的庞大规模令人望而却步。虽然常用的方法通过相同的体系结构提取或剪枝来减少PreLMs的大小，但我们探索将PreLMs提取为更有效的顺序感知嵌入模型。我们在GLUE基准测试中的结果表明，以嵌入为中心的学生（从BERT学习）在QQP和RTE上的得分与DistilBERT相当，通常与ELMo的得分相匹配或超过，并且只在检测语言可接受性方面落后。
摘要：Large-scale pretrained language models (PreLMs) are revolutionizing natural language processing across all benchmarks. However, their sheer size is prohibitive in low-resource or large-scale applications. While common approaches reduce the size of PreLMs via same-architecture distillation or pruning, we explore distilling PreLMs into more efficient order-aware embedding models. Our results on the GLUE benchmark show that embedding-centric students, which have learned from BERT, yield scores comparable to DistilBERT on QQP and RTE, often match or exceed the scores of ELMo, and only fall behind on detecting linguistic acceptability.
【5】 Learning Enhanced Optimisation for Routing Problems
标题：了解路由问题的增强型优化
链接：https://arxiv.org/abs/2109.08345
摘要：深度学习方法在解决路由问题方面已显示出良好的效果。然而，在机器学习算法和运筹学算法之间，在解决方案质量上仍然存在着巨大的差距。最近，又引入了另一条研究路线，它融合了机器学习和运筹学算法的优点。特别是，搜索微扰算子被用来改进解。然而，使用微扰可能无法保证高质量的解决方案。本文提出了“学习引导局部搜索”（L2GLS），这是一种基于学习的路由问题解决方法，使用惩罚项和强化学习自适应调整搜索努力。L2GLS将局部搜索（LS）算子的优势与惩罚项相结合，以逃避局部最优解。路由问题有许多实际应用，通常预先设定更大的实例，这对于学习优化领域中引入的许多现有算法来说仍然具有挑战性。我们证明了L2GLS在更大的TSP和CVRP上比其他机器学习方法取得了最新的结果。
摘要：Deep learning approaches have shown promising results in solving routing problems. However, there is still a substantial gap in solution quality between machine learning and operations research algorithms. Recently, another line of research has been introduced that fuses the strengths of machine learning and operational research algorithms. In particular, search perturbation operators have been used to improve the solution. Nevertheless, using the perturbation may not guarantee a quality solution. This paper presents "Learning to Guide Local Search" (L2GLS), a learning-based approach for routing problems that uses a penalty term and reinforcement learning to adaptively adjust search efforts. L2GLS combines local search (LS) operators' strengths with penalty terms to escape local optimals. Routing problems have many practical applications, often presetting larger instances that are still challenging for many existing algorithms introduced in the learning to optimise field. We show that L2GLS achieves the new state-of-the-art results on larger TSP and CVRP over other machine learning methods.
【6】 Decision Tree Learning with Spatial Modal Logics
标题：基于空间模态逻辑的决策树学习
链接：https://arxiv.org/abs/2109.08325
机构：Dept. of Mathematics and Computer Science, University of Ferrara, Italy, Dept. of Mathematical, Physical and Computer Sciences, University of Parma, Italy
备注：None
摘要：符号学习是解释性建模最直接的方法，但它的应用受到单一结构设计选择的阻碍：采用命题逻辑作为基础语言。最近，越来越多的命题符号学习方法开始出现，特别是对于时间相关数据。这些方法利用模式时态逻辑在强大的学习算法（如时态决策树）中的表达能力，其分类能力可与最佳非符号逻辑相媲美，同时生成具有显式知识表示的模型。为了在空间数据的情况下遵循相同的方法，本文中我们：i）提出了空间决策树学习的理论；ii）描述基于经典C4.5算法并严格扩展的空间决策树学习算法的原型实现；以及iii）进行了一系列实验，在这些实验中，我们比较了空间决策树与经典命题决策树在多个版本中对公共可用数据集上的多类图像分类问题的预测能力。我们的结果令人鼓舞，表明从命题模型到空间模型的性能有了明显的改善，从而显示出更高的可解释性水平。
摘要：Symbolic learning represents the most straightforward approach to interpretable modeling, but its applications have been hampered by a single structural design choice: the adoption of propositional logic as the underlying language. Recently, more-than-propositional symbolic learning methods have started to appear, in particular for time-dependent data. These methods exploit the expressive power of modal temporal logics in powerful learning algorithms, such as temporal decision trees, whose classification capabilities are comparable with the best non-symbolic ones, while producing models with explicit knowledge representation. With the intent of following the same approach in the case of spatial data, in this paper we: i) present a theory of spatial decision tree learning; ii) describe a prototypical implementation of a spatial decision tree learning algorithm based, and strictly extending, the classical C4.5 algorithm; and iii) perform a series of experiments in which we compare the predicting power of spatial decision trees with that of classical propositional decision trees in several versions, for a multi-class image classification problem, on publicly available datasets. Our results are encouraging, showing clear improvements in the performances from the propositional to the spatial models, which in turn show higher levels of interpretability.
【7】 Subtle Inverse Crimes: Naïvely training machine learning algorithms could lead to overly-optimistic results
标题：微妙的逆向犯罪：天真地训练机器学习算法可能会导致过于乐观的结果
链接：https://arxiv.org/abs/2109.08237
机构：UC Berkeley, UT Austin
备注：16 pages, 7 figures, two tables. Submitted to a journal
摘要 ：虽然开放式数据库在深度学习（DL）时代是一种重要的资源，但它们有时被“标示外”使用：为一项任务发布的数据用于为另一项任务训练算法。这项工作旨在强调，在某些情况下，这种常见做法可能会导致有偏见、过于乐观的结果。我们为反问题求解器演示了这一现象，并展示了隐藏数据预处理管道如何产生其偏差性能。我们描述了开放存取数据库的两个典型预处理管道，并研究了它们对三种成熟的磁共振成像（MRI）重建算法的影响：压缩感知（CS）、字典学习（DictL）和DL。在这项大规模的研究中，我们进行了大量的计算。我们的结果表明，CS、DictL和DL算法在对看似合适的数据（归一化均方根误差（NRMSE））进行训练时，会产生有系统偏差的结果随着预处理程度的不断提高，在某些情况下显示出25%-48%的人为增长。由于这种现象通常不为人所知，有偏见的结果有时会以最先进的方式发布；我们将其称为微妙的反向犯罪。因此，这项工作在na方面发出了红旗。”ive对大数据的标签外使用，揭示了现代反问题解决者对由此产生的偏见的脆弱性。
摘要：While open databases are an important resource in the Deep Learning (DL) era, they are sometimes used "off-label": data published for one task are used for training algorithms for a different one. This work aims to highlight that in some cases, this common practice may lead to biased, overly-optimistic results. We demonstrate this phenomenon for inverse problem solvers and show how their biased performance stems from hidden data preprocessing pipelines. We describe two preprocessing pipelines typical of open-access databases and study their effects on three well-established algorithms developed for Magnetic Resonance Imaging (MRI) reconstruction: Compressed Sensing (CS), Dictionary Learning (DictL), and DL. In this large-scale study we performed extensive computations. Our results demonstrate that the CS, DictL and DL algorithms yield systematically biased results when na\"ively trained on seemingly-appropriate data: the Normalized Root Mean Square Error (NRMSE) improves consistently with the preprocessing extent, showing an artificial increase of 25%-48% in some cases. Since this phenomenon is generally unknown, biased results are sometimes published as state-of-the-art; we refer to that as subtle inverse crimes. This work hence raises a red flag regarding na\"ive off-label usage of Big Data and reveals the vulnerability of modern inverse problem solvers to the resulting bias.
【8】 Deep Spiking Neural Networks with Resonate-and-Fire Neurons
标题：具有共振和放电神经元的深度尖峰神经网络
链接：https://arxiv.org/abs/2109.08234
机构：Independent, Microsoft EGDC, D. Bernal-Casas, U Barcelona, Spain
备注：Preprint
摘要：在这项工作中，我们探索了一种新的尖峰神经网络（SNN）公式，其中共振和激发（RAF）神经元（Izhikevich，2001）通过反向传播进行梯度下降训练。RAF-SNN虽然在生物学上更合理，但在不同的网络配置中，使用相似或更少的参数，其性能可与机器学习文献中的传统模型相比或更高。引人注目的是，RAF-SNN在静态和动态条件下，对测试/训练时产生的噪声具有鲁棒性。与MNIST上的CNN相比，我们在测试时显示N（0，0.2）诱导噪声的绝对准确度高出25%。与N-MNIST上的LSTM相比，我们在训练时显示了70%的绝对准确度和20%的诱导噪声。
摘要：In this work, we explore a new Spiking Neural Network (SNN) formulation with Resonate-and-Fire (RAF) neurons (Izhikevich, 2001) trained with gradient descent via back-propagation. The RAF-SNN, while more biologically plausible, achieves performance comparable to or higher than conventional models in the Machine Learning literature across different network configurations, using similar or fewer parameters. Strikingly, the RAF-SNN proves robust against noise induced at testing/training time, under both static and dynamic conditions. Against CNN on MNIST, we show 25% higher absolute accuracy with N(0, 0.2) induced noise at testing time. Against LSTM on N-MNIST, we show 70% higher absolute accuracy with 20% induced noise at training time.
【9】 SLAW: Scaled Loss Approximate Weighting for Efficient Multi-Task Learning
标题：一种高效多任务学习的缩放损失近似加权算法
链接：https://arxiv.org/abs/2109.08218
机构：George Mason University
摘要：多任务学习（Multi-task learning，MTL）是机器学习的一个分支，有着重要的应用，但MTL的多目标优化特性导致任务间训练难以平衡。最好的MTL优化方法需要单独计算每个任务的损失函数的梯度，这阻碍了对大量任务的可伸缩性。在本文中，我们提出了标度损失近似加权（SLAW），这是一种用于多任务优化的方法，它与现有最佳方法的性能相匹配，同时效率更高。SLAW通过估计每个任务梯度的大小来平衡任务之间的学习，而不执行任何额外的向后传递。我们为SLAW对梯度量级的估计提供了理论和经验依据。非线性回归、多任务计算机视觉和药物发现虚拟筛选的实验结果表明，SLAW在不牺牲性能的情况下比强基线更有效，适用于多种领域。
摘要：Multi-task learning (MTL) is a subfield of machine learning with important applications, but the multi-objective nature of optimization in MTL leads to difficulties in balancing training between tasks. The best MTL optimization methods require individually computing the gradient of each task's loss function, which impedes scalability to a large number of tasks. In this paper, we propose Scaled Loss Approximate Weighting (SLAW), a method for multi-task optimization that matches the performance of the best existing methods while being much more efficient. SLAW balances learning between tasks by estimating the magnitudes of each task's gradient without performing any extra backward passes. We provide theoretical and empirical justification for SLAW's estimation of gradient magnitudes. Experimental results on non-linear regression, multi-task computer vision, and virtual screening for drug discovery demonstrate that SLAW is significantly more efficient than strong baselines without sacrificing performance and applicable to a diverse range of domains.
其他(13篇)
【1】 Realistic PointGoal Navigation via Auxiliary Losses and Information Bottleneck
标题：辅助性损失和信息瓶颈下的现实点目标导航
链接：https://arxiv.org/abs/2109.08677
机构： Georgia Institute of Technology, Atlanta, GA, Facebook AI Research, Menlo Park, CA
摘要：我们提出了一种新的用于训练真实点目标导航的体系结构和训练范式——在驱动和传感器噪声下在看不见的环境中导航到目标坐标，而无需进行地面真实定位。具体地说，我们发现，在这种环境下，主要的挑战是学习本地化——当去除理想化的本地化时，代理无法准确地停止在目标上，尽管在实现目标方面取得了可靠的进展。为了解决这个问题，我们引入了一组辅助损失来帮助代理学习本地化。此外，我们还探讨了将代理的精确位置视为特权信息的想法——它在测试时不可用，但在模拟训练时可用。我们通过一个信息瓶颈，允许代理在训练期间限制访问地面实况本地化读数。在此设置下，代理将因使用此特权信息而受到惩罚，鼓励代理仅在对学习至关重要的情况下利用此信息。这使代理能够首先学习导航，然后学习本地化，而不是在训练中混淆这两个目标。我们在半理想（无噪声模拟，无罗盘+GPS）和真实（添加噪声模拟）设置中评估了我们提出的方法。具体而言，我们的方法在半理想化设置下的性能优于现有基线，在现实设置下的性能优于现有基线，分别为18\%/21\%SPL/成功和15\%/20\%SPL。我们改进的成功率和SPL指标表明，我们的代理在保持强大导航策略的同时，提高了准确自我定位的能力。我们的实现可以在https://github.com/NicoGrande/habitat-pointnav-via-ib.
摘要：We propose a novel architecture and training paradigm for training realistic PointGoal Navigation -- navigating to a target coordinate in an unseen environment under actuation and sensor noise without access to ground-truth localization. Specifically, we find that the primary challenge under this setting is learning localization -- when stripped of idealized localization, agents fail to stop precisely at the goal despite reliably making progress towards it. To address this we introduce a set of auxiliary losses to help the agent learn localization. Further, we explore the idea of treating the precise location of the agent as privileged information -- it is unavailable during test time, however, it is available during training time in simulation. We grant the agent restricted access to ground-truth localization readings during training via an information bottleneck. Under this setting, the agent incurs a penalty for using this privileged information, encouraging the agent to only leverage this information when it is crucial to learning. This enables the agent to first learn navigation and then learn localization instead of conflating these two objectives in training. We evaluate our proposed method both in a semi-idealized (noiseless simulation without Compass+GPS) and realistic (addition of noisy simulation) settings. Specifically, our method outperforms existing baselines on the semi-idealized setting by 18\%/21\% SPL/Success and by 15\%/20\% SPL in the realistic setting. Our improved Success and SPL metrics indicate our agent's improved ability to accurately self-localize while maintaining a strong navigation policy. Our implementation can be found at https://github.com/NicoGrande/habitat-pointnav-via-ib.
【2】 Is Curiosity All You Need? On the Utility of Emergent Behaviours from Curious Exploration
标题：你只需要好奇心吗？从好奇探索看突发行为的效用
链接：https://arxiv.org/abs/2109.08603
机构：DeepMind,University of Oxford,Queensland University of Technology
备注：14 pages, 7 figures, 2 tables
摘要 ：基于好奇心的奖励机制可以提供强大的探索机制，有助于发现复杂、稀疏或长期任务的解决方案。然而，当代理学会到达以前未开发的空间，并且目标适应奖励新区域时，许多行为出现，但由于被不断变化的目标覆盖而消失。我们认为，仅仅将好奇心用于快速环境探索或作为对特定任务的奖励，并不能充分利用这项技术的潜力，也会错过有用的技能。相反，我们建议将重点转向保留基于好奇心的学习过程中出现的行为。我们认为，这些自我发现的行为是特工解决相关任务的重要技能。我们的实验证明了在整个训练过程中行为的不断变化，以及简单的策略快照方法对传输任务重用发现的行为的好处。
摘要：Curiosity-based reward schemes can present powerful exploration mechanisms which facilitate the discovery of solutions for complex, sparse or long-horizon tasks. However, as the agent learns to reach previously unexplored spaces and the objective adapts to reward new areas, many behaviours emerge only to disappear due to being overwritten by the constantly shifting objective. We argue that merely using curiosity for fast environment exploration or as a bonus reward for a specific task does not harness the full potential of this technique and misses useful skills. Instead, we propose to shift the focus towards retaining the behaviours which emerge during curiosity-based learning. We posit that these self-discovered behaviours serve as valuable skills in an agent's repertoire to solve related tasks. Our experiments demonstrate the continuous shift in behaviour throughout training and the benefits of a simple policy snapshot method to reuse discovered behaviour for transfer tasks.
【3】 Measuring Fairness under Unawareness via Quantification
标题：不知不觉中的公正性量化测量
链接：https://arxiv.org/abs/2109.08549
机构：Received: date Accepted: date
摘要：通过监督学习训练的模型越来越多地应用于高风险领域，当它们的预测为人们的决策提供信息时，它们不可避免地会对他们的生活产生（积极或消极）影响。因此，负责开发这些模型的人必须仔细评估它们对不同人群的影响，并确保敏感的人口特征，如种族或性别，不会导致对特定群体成员的不公平待遇。要做到这一点，评估模型影响的人员必须了解人口特征。不幸的是，这些属性的收集常常与行业惯例和关于数据最小化和隐私的立法相冲突。因此，即使是在开发模型的公司内部，也很难衡量经过训练的模型的群体公平性。在这项工作中，我们通过使用量化技术来解决在不了解敏感属性的情况下测量群体公平性的问题，量化是一项监督学习任务，直接提供群体水平的患病率估计（而不是个人水平的类别标签）。我们确定了在不知情情况下使公平性估计复杂化的五个重要因素，并将它们形式化为五个不同的实验协议，在此协议下，我们评估了群体公平性不同估计量的有效性。我们还考虑潜在模型误用的问题，以推断在个体层面上的敏感属性，并证明量化方法适合去耦（期望的）测量群体公平性的目标（不期望的）推断个体敏感属性的目标。
摘要：Models trained by means of supervised learning are increasingly deployed in high-stakes domains, and, when their predictions inform decisions about people, they inevitably impact (positively or negatively) on their lives. As a consequence, those in charge of developing these models must carefully evaluate their impact on different groups of people and ensure that sensitive demographic attributes, such as race or sex, do not result in unfair treatment for members of specific groups. For doing this, awareness of demographic attributes on the part of those evaluating model impacts is fundamental. Unfortunately, the collection of these attributes is often in conflict with industry practices and legislation on data minimization and privacy. For this reason, it may be hard to measure the group fairness of trained models, even from within the companies developing them. In this work, we tackle the problem of measuring group fairness under unawareness of sensitive attributes, by using techniques from quantification, a supervised learning task concerned with directly providing group-level prevalence estimates (rather than individual-level class labels). We identify five important factors that complicate the estimation of fairness under unawareness and formalize them into five different experimental protocols under which we assess the effectiveness of different estimators of group fairness. We also consider the problem of potential model misuse to infer sensitive attributes at an individual level, and demonstrate that quantification approaches are suitable for decoupling the (desirable) objective of measuring group fairness from the (undesirable) objective of inferring sensitive attributes of individuals.
【4】 Soft Actor-Critic With Integer Actions
标题：软演员--整体式行动的批评家
链接：https://arxiv.org/abs/2109.08512
机构： Princeton University, Siemens Corporate Technology
摘要：强化学习在离散动作下得到了很好的研究。Integer actions设置在业界很受欢迎，但由于其高维性，仍然具有挑战性。为此，我们研究了整数动作下的强化学习，将软参与者批评（SAC）算法与整数重参数化相结合。我们对整型动作的关键观察是，可以使用其相似性属性简化其离散结构。因此，所提出的整数重参数化不需要一个热编码，并且是低维的。实验表明，在整数动作下，该算法在机器人控制任务上的性能与连续动作算法相当，在配电系统控制任务上的性能优于近端策略优化算法。
摘要：Reinforcement learning is well-studied under discrete actions. Integer actions setting is popular in the industry yet still challenging due to its high dimensionality. To this end, we study reinforcement learning under integer actions by incorporating the Soft Actor-Critic (SAC) algorithm with an integer reparameterization. Our key observation for integer actions is that their discrete structure can be simplified using their comparability property. Hence, the proposed integer reparameterization does not need one-hot encoding and is of low dimensionality. Experiments show that the proposed SAC under integer actions is as good as the continuous action version on robot control tasks and outperforms Proximal Policy Optimization on power distribution systems control tasks.
【5】 SaCoFa: Semantics-aware Control-flow Anonymization for Process Mining
标题：SaCoFa：面向进程挖掘的语义感知控制流匿名化
链接：https://arxiv.org/abs/2109.08501
机构：Fabian R¨osel∗, ∗Humboldt-Universit¨at zu Berlin, Berlin, Germany, †University of Mannheim, Mannheim, Germany
摘要：保护隐私的流程挖掘支持使用事件日志分析业务流程，同时保证保护流程涉众的敏感信息。为此，现有方法会对提取事件日志属性（如跟踪变量的频率分布）以进行分析的查询结果添加噪声。但是，噪声插入忽略了过程的语义，可能会生成原始日志中不存在的跟踪。这是有问题的。它降低了已发布数据的效用，并使噪音易于识别，因为一些痕迹会违反众所周知的语义约束。因此，在本文中，我们主张采用过程语义来保护隐私。对于常见的跟踪变量查询，我们展示了如何基于指数机制，引入语义约束以确保查询结果的差异保密性。实验表明，我们的语义感知匿名化方法产生的事件日志比现有方法具有更高的实用性。
摘要：Privacy-preserving process mining enables the analysis of business processes using event logs, while giving guarantees on the protection of sensitive information on process stakeholders. To this end, existing approaches add noise to the results of queries that extract properties of an event log, such as the frequency distribution of trace variants, for analysis.Noise insertion neglects the semantics of the process, though, and may generate traces not present in the original log. This is problematic. It lowers the utility of the published data and makes noise easily identifiable, as some traces will violate well-known semantic constraints.In this paper, we therefore argue for privacy preservation that incorporates a process semantics. For common trace-variant queries, we show how, based on the exponential mechanism, semantic constraints are incorporated to ensure differential privacy of the query result. Experiments demonstrate that our semantics-aware anonymization yields event logs of significantly higher utility than existing approaches.
【6】 Accurate, Interpretable, and Fast Animation: AnIterative, Sparse, and Nonconvex Approach
标题：精确、可解释和快速的动画：迭代、稀疏和非凸性方法
链接：https://arxiv.org/abs/2109.08356
机构： Jakoveti´c is with the Department of Mathematics
摘要：数字人体动画依赖于高质量的人脸3D模型：rigs。面钻机必须精确，同时计算速度也很快。最常见的索具模型之一是blendshape模型。我们提出了一种新的算法来解决人脸动画中的非凸逆装配问题。我们的方法是基于模型的，但与以前基于模型的方法相比，我们对高阶钻机模型使用二次近似而不是线性近似。这平均将解决方案的精度提高了8%，并且，经经验结果证实，增加了结果参数向量的稀疏性——这是动画艺术家可解释性的一个重要特征。所提出的解决方案基于Levenberg-Marquardt（LM）算法，应用于具有稀疏正则化的非凸约束问题。为了降低迭代的复杂性，进一步调用了优化最小化（MM）范式，这导致了一个易于解决的问题，该问题在每个算法迭代中的参数是可分离的。该算法在大量专有和开源的动画数据集上进行了评估，结果表明，与基于线性钻机近似的标准方法相比，我们的方法具有优越性。虽然我们的算法针对特定的问题，但它可能有额外的信号处理应用。
摘要 ：Digital human animation relies on high-quality 3D models of the human face: rigs. A face rig must be accurate and, at the same time, fast to compute. One of the most common rigging models is the blendshape model. We propose a novel algorithm for solving the nonconvex inverse rig problem in facial animation. Our approach is model-based, but in contrast with previous model-based approaches, we use a quadratic instead of the linear approximation to the higher order rig model. This increases the accuracy of the solution by 8 percent on average and, confirmed by the empirical results, increases the sparsity of the resulting parameter vector -- an important feature for interpretability by animation artists. The proposed solution is based on a Levenberg-Marquardt (LM) algorithm, applied to a nonconvex constrained problem with sparsity regularization. In order to reduce the complexity of the iterates, a paradigm of Majorization Minimization (MM) is further invoked, which leads to an easy to solve problem that is separable in the parameters at each algorithm iteration. The algorithm is evaluated on a number of animation datasets, proprietary and open-source, and the results indicate the superiority of our method compared to the standard approach based on the linear rig approximation. Although our algorithm targets the specific problem, it might have additional signal processing applications.
【7】 Dropout's Dream Land: Generalization from Learned Simulators to Reality
标题：辍学的梦境：从学识模拟器到现实的概括
链接：https://arxiv.org/abs/2109.08342
机构：Department of Computer Science and Engineering, Hong Kong University of Science and Technology, Hong Kong
备注：Published at ECML PKDD 2021
摘要：世界模型是用于模拟环境的生成模型。事实证明，世界模型能够学习强化学习环境的空间和时间表示。在某些情况下，世界模型为代理提供了完全在自己的梦想环境中学习的机会。在这项工作中，我们探索如何提高从梦境环境到真实环境（Dream2Real）的泛化能力。我们提出了一种通用的方法，以提高控制器的能力，从一个神经网络的梦想环境转移到现实中，几乎没有额外的成本。这些改进是从领域随机化中获得的灵感，其基本思想是在不从根本上改变手头任务的情况下尽可能多地随机化模拟器。通常，域随机化假设访问具有可配置参数的预构建模拟器，但通常不可用。通过使用辍学训练世界模型，梦想环境能够创建几乎无限多个不同的梦想环境。以前的辍学用例在推断时不使用辍学，或者对多个抽样掩码生成的预测进行平均（蒙特卡罗辍学）。辍学者的梦想之地利用每一个独特的面具来创造一个多样化的梦想环境。我们的实验结果表明，辍学者的梦境是弥合梦境与现实之间的现实鸿沟的有效技术。此外，我们还进行了一系列广泛的消融研究。
摘要：A World Model is a generative model used to simulate an environment. World Models have proven capable of learning spatial and temporal representations of Reinforcement Learning environments. In some cases, a World Model offers an agent the opportunity to learn entirely inside of its own dream environment. In this work we explore improving the generalization capabilities from dream environments to real environments (Dream2Real). We present a general approach to improve a controller's ability to transfer from a neural network dream environment to reality at little additional cost. These improvements are gained by drawing on inspiration from Domain Randomization, where the basic idea is to randomize as much of a simulator as possible without fundamentally changing the task at hand. Generally, Domain Randomization assumes access to a pre-built simulator with configurable parameters but oftentimes this is not available. By training the World Model using dropout, the dream environment is capable of creating a nearly infinite number of different dream environments. Previous use cases of dropout either do not use dropout at inference time or averages the predictions generated by multiple sampled masks (Monte-Carlo Dropout). Dropout's Dream Land leverages each unique mask to create a diverse set of dream environments. Our experimental results show that Dropout's Dream Land is an effective technique to bridge the reality gap between dream environments and reality. Furthermore, we additionally perform an extensive set of ablation studies.
【8】 A Logic-based Multi-agent System for Ethical Monitoring and Evaluation of Dialogues
标题：一种基于逻辑的对话伦理监控与评估多智能体系统
链接：https://arxiv.org/abs/2109.08294
机构：DISIM, University of L’Aquila, Italy, DIB & CILA, University of Bari ”Aldo Moro”, Italy
备注：None
摘要：对话系统是为人机交互的各种实际目的而设计的工具。这些系统应该建立在道德基础上，因为它们的行为可能会严重影响用户（尤其是儿童）。本文的主要目的是介绍一个多智能体系统（MAS）的体系结构和原型实现，该系统用于对话系统的道德监控和评估。开发并展示了一个原型应用程序，用于监控和评估在线客户服务聊天点中的聊天代理（人/人工）道德行为，即其机构/公司的道德和行为准则。讨论了本研究的未来工作和有待解决的问题。
摘要：Dialogue Systems are tools designed for various practical purposes concerning human-machine interaction. These systems should be built on ethical foundations because their behavior may heavily influence a user (think especially about children). The primary objective of this paper is to present the architecture and prototype implementation of a Multi Agent System (MAS) designed for ethical monitoring and evaluation of a dialogue system. A prototype application, for monitoring and evaluation of chatting agents' (human/artificial) ethical behavior in an online customer service chat point w.r.t their institution/company's codes of ethics and conduct, is developed and presented. Future work and open issues with this research are discussed.
【9】 Strategic Ranking
标题：战略排名
链接：https://arxiv.org/abs/2109.08240
机构：∗Department of Electrical Engineering and Computer Sciences, University of California, Berkeley, †School of Operations Research and Information Engineering, Cornell Tech and the Technion
备注：35 pages, 3 figures
摘要：策略分类研究对策略个体输入操作具有鲁棒性的分类器的设计。然而，现有文献并未考虑算法设计引起的个体间竞争的影响。受大学入学等受限分配设置的激励，我们引入了战略排名，其中（设计的）个人奖励取决于申请人在兴趣衡量中的努力后排名。我们的结果说明了申请者之间的竞争如何影响最终的均衡和模型洞察。我们分析了各种排名奖励设计是如何权衡申请人、学校和社会效用的，特别是排名设计是如何应对因不同的资源获取方式而产生的不平等，从而提高一个人的测量分数的：我们发现排名奖励设计中的随机化可以缓解两种不同影响的措施，福利差距和获得机会，而非随机化可能导致高水平的竞争，系统地排斥弱势群体。
摘要：Strategic classification studies the design of a classifier robust to the manipulation of input by strategic individuals. However, the existing literature does not consider the effect of competition among individuals as induced by the algorithm design. Motivated by constrained allocation settings such as college admissions, we introduce strategic ranking, in which the (designed) individual reward depends on an applicant's post-effort rank in a measurement of interest. Our results illustrate how competition among applicants affects the resulting equilibria and model insights. We analyze how various ranking reward designs trade off applicant, school, and societal utility and in particular how ranking design can counter inequities arising from disparate access to resources to improve one's measured score: We find that randomization in the ranking reward design can mitigate two measures of disparate impact, welfare gap and access, whereas non-randomization may induce a high level of competition that systematically excludes a disadvantaged group.
【10】 KATANA: Simple Post-Training Robustness Using Test Time Augmentations
标题：Katana：使用测试时间增加的简单训练后健壮性
链接：https://arxiv.org/abs/2109.08191
机构：Department of Electrical Engineering, Tel Aviv University, Tel Aviv, Israel
摘要：尽管深度神经网络（DNN）在许多实际任务中取得了优异的性能，但它们极易受到敌对攻击。对抗此类攻击的主要防御措施是对抗性训练，在这种技术中，DNN通过在其输入中引入对抗性噪声来训练对对抗性攻击的鲁棒性。该程序有效，但必须在训练阶段完成。在这项工作中，我们提出了一种新的简单易用的技术，KATANA，用于在不修改其权重的情况下对现有预训练DNN进行鲁棒化。对于每幅图像，我们通过应用不同的颜色、模糊、噪声和几何变换生成N个随机测试时间增强（TTA）。接下来，我们利用DNN的logits输出来训练一个简单的随机森林分类器来预测真实的类标签。我们的策略实现了对各种攻击的最先进的对抗鲁棒性，同时对自然图像的分类进行了最小程度的妥协。我们还针对两种自适应白盒攻击对KATANA进行了测试，并将其与对抗性训练相结合，显示出优异的效果。代码在中提供https://github.com/giladcohen/KATANA.
摘要 ：Although Deep Neural Networks (DNNs) achieve excellent performance on many real-world tasks, they are highly vulnerable to adversarial attacks. A leading defense against such attacks is adversarial training, a technique in which a DNN is trained to be robust to adversarial attacks by introducing adversarial noise to its input. This procedure is effective but must be done during the training phase. In this work, we propose a new simple and easy-to-use technique, KATANA, for robustifying an existing pretrained DNN without modifying its weights. For every image, we generate N randomized Test Time Augmentations (TTAs) by applying diverse color, blur, noise, and geometric transforms. Next, we utilize the DNN's logits output to train a simple random forest classifier to predict the real class label. Our strategy achieves state-of-the-art adversarial robustness on diverse attacks with minimal compromise on the natural images' classification. We test KATANA also against two adaptive white-box attacks and it shows excellent results when combined with adversarial training. Code is available in https://github.com/giladcohen/KATANA.
【11】 Sparse Factorization of Large Square Matrices
标题：大方阵的稀疏分解
链接：https://arxiv.org/abs/2109.08184
机构：Norwegian University of Science and Technology
摘要：方阵出现在许多机器学习问题和模型中。在大型方阵上进行优化在内存和时间上都很昂贵。因此，需要一个经济近似值。传统的近似方法将平方矩阵分解为许多低阶矩阵。然而，如果近似矩阵本质上是高秩或接近满秩，则低秩约束是性能瓶颈。在本文中，我们提出用稀疏满秩矩阵的乘积来近似一个大的平方矩阵。在近似情况下，我们的方法只需要$N（\log N）^2$个非零数字，就可以得到一个$N\乘以N$的完整矩阵。我们提出了非参数和参数两种方法来寻找因式分解。前者直接学习分解矩阵，后者训练神经网络将输入数据映射到非零矩阵项。稀疏因子分解方法在各种合成和真实的方阵中进行了测试。实验结果表明，当近似矩阵稀疏且秩较高时，该方法具有较好的逼近性能。基于这一发现，我们使用我们的参数化方法作为一种可伸缩的注意结构，在长序列数据的学习任务中表现出色，并击败Transformer及其几个变体。
摘要：Square matrices appear in many machine learning problems and models. Optimization over a large square matrix is expensive in memory and in time. Therefore an economic approximation is needed. Conventional approximation approaches factorize the square matrix into a number matrices of much lower ranks. However, the low-rank constraint is a performance bottleneck if the approximated matrix is intrinsically high-rank or close to full rank. In this paper, we propose to approximate a large square matrix with a product of sparse full-rank matrices. In the approximation, our method needs only $N(\log N)^2$ non-zero numbers for an $N\times N$ full matrix. We present both non-parametric and parametric ways to find the factorization. In the former, we learn the factorizing matrices directly, and in the latter, we train neural networks to map input data to the non-zero matrix entries. The sparse factorization method is tested for a variety of synthetic and real-world square matrices. The experimental results demonstrate that our method gives a better approximation when the approximated matrix is sparse and high-rank. Based on this finding, we use our parametric method as a scalable attention architecture that performs strongly in learning tasks for long sequential data and defeats Transformer and its several variants.
【12】 Interpretable Local Tree Surrogate Policies
标题：可解释的本地树代理策略
链接：https://arxiv.org/abs/2109.08180
机构： Department of Aeronautics and Astronautics, Stanford University, Department of Computer Science, Stanford University
备注：pre-print, submitted to AAAI 2022 Conference, 7 pages
摘要：高维政策，如神经网络所代表的政策，人类无法合理解释。这种可解释性的缺乏降低了用户对策略行为的信任，限制了他们对低影响任务（如视频游戏）的使用。不幸的是，许多方法依赖于神经网络表示来进行有效的学习。在这项工作中，我们提出了一种方法来建立可预测的策略树作为策略（如神经网络）的代理。政策树很容易被人理解，并提供未来行为的定量预测。我们在几个模拟任务上演示了这种方法的性能。
摘要：High-dimensional policies, such as those represented by neural networks, cannot be reasonably interpreted by humans. This lack of interpretability reduces the trust users have in policy behavior, limiting their use to low-impact tasks such as video games. Unfortunately, many methods rely on neural network representations for effective learning. In this work, we propose a method to build predictable policy trees as surrogates for policies such as neural networks. The policy trees are easily human interpretable and provide quantitative predictions of future behavior. We demonstrate the performance of this approach on several simulated tasks.
【13】 A Cautionary Tale of Decorrelating Theory Uncertainties
标题：解相关理论不确定性的警示故事
链接：https://arxiv.org/abs/2109.08159
机构：Department of Physics and Astronomy, University of California, Irvine, CA , USA, Physics Division, Lawrence Berkeley National Laboratory, Berkeley, CA , USA, Berkeley Institute for Data Science, University of California, Berkeley, CA , USA
备注：19 pages, 8 figures
摘要：人们提出了多种技术来训练独立于给定特征的机器学习分类器。虽然这可能是实现背景估计的基本技术，但也可能有助于减少不确定性。我们仔细研究了理论上的不确定性，这些不确定性通常没有统计来源。我们将提供两点（碎片建模）和连续（高阶校正）不确定性的明确示例，其中解相关显著降低了表观不确定性，而实际不确定性则大得多。这些结果表明，在对这些类型的不确定性使用去相关时，只要我们没有将其完全分解为具有统计意义的分量，就应该谨慎。
摘要：A variety of techniques have been proposed to train machine learning classifiers that are independent of a given feature. While this can be an essential technique for enabling background estimation, it may also be useful for reducing uncertainties. We carefully examine theory uncertainties, which typically do not have a statistical origin. We will provide explicit examples of two-point (fragmentation modeling) and continuous (higher-order corrections) uncertainties where decorrelating significantly reduces the apparent uncertainty while the actual uncertainty is much larger. These results suggest that caution should be taken when using decorrelation for these types of uncertainties as long as we do not have a complete decomposition into statistically meaningful components.
机器翻译，仅供参考
点击“阅读原文”获取带摘要的学术速递
Python社区是高质量的Python/Django开发社区
本文地址：http://www.python88.com/topic/120301
本文地址：http://www.python88.com/topic/120301
497 次点击