Update！H5支撑摘要折叠，体会更佳！点击拜访arxivdaily.com
cs.LG 方向，今天合计64篇
Graph相关(图学习|图神经网络|图优化等)(2篇)
【1】 Label-informed Graph Structure Learning for Node Classification
标题：依据标签信息的图结构学习在节点分类中的运用
链接：https://arxiv.org/abs/2108.04595
安排：Center for Research on Intelligent Perception and Computing, Institute of Automation, Chinese Academy of Sciences, School of Artificial Intelligence, University of Chinese Academy of Sciences
补白：Accepted by CIKM 2021 short paper track
摘要：图形神经网络（GNNs）在各个范畴都取得了巨大的成功。但是，大多数GNN办法对图结构的质量十分灵敏。为了处理这个问题，一些研讨运用不同的图结构学习战略来细化原始图结构。但是，这些办法只考虑特征信息，而疏忽可用的标签信息。在本文中，咱们提出了一种新的标签告诉图结构学习结构，该结构经过类搬运矩阵显式地结合标签信息。咱们在七节点分类基准数据集进步行了很多试验，成果标明咱们的办法优于或匹配最新的基线。
摘要：Graph Neural Networks (GNNs) have achieved great success among various
domains. Nevertheless, most GNN methods are sensitive to the quality of graph
structures. To tackle this problem, some studies exploit different graph
structure learning strategies to refine the original graph structure. However,
these methods only consider feature information while ignoring available label
information. In this paper, we propose a novel label-informed graph structure
learning framework which incorporates label information explicitly through a
class transition matrix. We conduct extensive experiments on seven node
classification benchmark datasets and the results show that our method
outperforms or matches the state-of-the-art baselines.
【2】 Learning Multi-Granular Spatio-Temporal Graph Network for Skeleton-based Action Recognition
标题：依据骨架的动作辨认多粒度时空图网络学习
链接：https://arxiv.org/abs/2108.04536
安排：Open Lab, Newcastle University, Newcastle upon Tyne, UK, Department of Computer Vision Technology (VIS), Baidu Inc., China, ShanghaiTech University, Shanghai, China
补白：Accepted by ACM MM'21
摘要：因为人体运动的多粒度和大变异性，依据骨架的动作辨认使命依然是以人为中心的场景了解的中心应战。现有的办法一般对不同的运动形式运用单一的神经标明，在有限的练习数据下难以捕获细粒度的动作类。为了处理上述问题，咱们提出了一种用于依据骨架的动作分类的多粒度时空图网络，该网络联合建模了粗粒度和细粒度的骨架运动形式。为此，咱们开发了一个由两个穿插分支组成的双头图网络，它使咱们能够以高效的办法在两个时空分辨率下提取特征。此外，咱们的网络选用跨部分交流战略，以彼此增强两个部分主管的代表性。咱们在三个大型数据集，即NTU RGB+D 60、NTU RGB+D 120和动力学骨架进步行了广泛的试验，并在一切基准上完结了最先进的功能，这验证了咱们办法的有用性。
摘要
GAN|对立|进犯|生成相关(9篇)
【1】 Continual Learning for Grounded Instruction Generation by Observing Human Following Behavior
标题：经过查询人的跟从行为完结扎根指令生成的继续学习
链接：https://arxiv.org/abs/2108.04812
安排：Department of Computer Science and Cornell Tech, Cornell University
补白
摘要
摘要：We study continual learning for natural language instruction generation, by
observing human users' instruction execution. We focus on a collaborative
scenario, where the system both acts and delegates tasks to human users using
natural language. We compare user execution of generated instructions to the
original system intent as an indication to the system's success communicating
its intent. We show how to use this signal to improve the system's ability to
generate instructions via contextual bandit learning. In interaction with real
users, our system demonstrates dramatic improvements in its ability to generate
language over time.
【2】 Correlation Clustering Reconstruction in Semi-Adversarial Models
标题：半对立模型中的相关聚类重构
链接：https://arxiv.org/abs/2108.04729
安排：Sapienza University, Rome, Italy, Bocconi University, Milan, Italy
摘要
摘要：Correlation Clustering is an important clustering problem with many
applications. We study the reconstruction version of this problem in which one
is seeking to reconstruct a latent clustering that has been corrupted by random
noise and adversarial modifications.
Concerning the latter, we study a standard "post-adversarial" model, in which
adversarial modifications come after the noise, and also introduce and analyze
a "pre-adversarial" model in which adversarial modifications come before the
noise. Given an input coming from such a semi-adversarial generative model, the
goal is to reconstruct almost perfectly and with high probability the latent
clustering.
We focus on the case where the hidden clusters have equal size and show the
following. In the pre-adversarial setting, spectral algorithms are optimal, in
the sense that they reconstruct all the way to the information-theoretic
threshold beyond which no reconstruction is possible. In contrast, in the
post-adversarial setting their ability to restore the hidden clusters stops
before the threshold, but the gap is optimally filled by SDP-based algorithms.
【3】 UniNet: A Unified Scene Understanding Network and Exploring Multi-Task Relationships through the Lens of Adversarial Attacks
标题：UniNet：一个一致的场景了解网络，经过对立性进犯的镜头探究多使命联系
链接：https://arxiv.org/abs/2108.04584
安排：Advanced Research Lab, NavInfo Europe, Eindhoven, The Netherlands
补白：Accepted at DeepMTL workshop, ICCV 2021
摘要：场景了解关于打算在实在世界中运转的自治体系来说至关重要。单使命视觉网络仅依据场景的某些方面提取信息。另一方面，在多使命学习（MTL）中，这些单一使命是联合学习的，然后为使命同享信息和取得更全面的了解供给了时机。为此，咱们开发了UniNet，这是一个一致的场景了解网络，能够精确有用地揣度重要的视觉使命，包含方针检测、语义切割、实例切割、单目深度估量和单目实例深度猜测。当这些使命查看不同的语义和几许信息时，它们能够彼此弥补，也能够彼此抵触。因而，了解使命间的联系能够供给有用的头绪，以完结互补信息同享。咱们经过对立性进犯的视角来点评UniNet中的使命联系，依据这样一个概念，即它们能够运用神经网络中的学习差错和使命交互。在Cityscapes数据集进步行的很多试验（运用非方针进犯和方针进犯）标明，语义使命之间彼此作用激烈，几许使命也是如此。此外，咱们还发现，语义使命和几许使命之间的联系是不对称的，跟着咱们向更高层次的表达办法开展，它们之间的交互作用变得越来越弱。
摘要：Scene understanding is crucial for autonomous systems which intend to operate
in the real world. Single task vision networks extract information only based
on some aspects of the scene. In multi-task learning (MTL), on the other hand,
these single tasks are jointly learned, thereby providing an opportunity for
tasks to share information and obtain a more comprehensive understanding. To
this end, we develop UniNet, a unified scene understanding network that
accurately and efficiently infers vital vision tasks including object
detection, semantic segmentation, instance segmentation, monocular depth
estimation, and monocular instance depth prediction. As these tasks look at
different semantic and geometric information, they can either complement or
conflict with each other. Therefore, understanding inter-task relationships can
provide useful cues to enable complementary information sharing. We evaluate
the task relationships in UniNet through the lens of adversarial attacks based
on the notion that they can exploit learned biases and task interactions in the
neural network. Extensive experiments on the Cityscapes dataset, using
untargeted and targeted attacks reveal that semantic tasks strongly interact
amongst themselves, and the same holds for geometric tasks. Additionally, we
show that the relationship between semantic and geometric tasks is asymmetric
and their interaction becomes weaker as we move towards higher-level
representations.
【4】 Regularized Sequential Latent Variable Models with Adversarial Neural Networks
标题：依据对立性神经网络的正则化序贯潜变量模型
链接：https://arxiv.org/abs/2108.04496
安排：the Department of Information Science and Engineering, KTH Royal Institute of Technology, Stockholm, Sweden
补白：A-VRNN
摘要：递归神经网络（RNN）具有丰厚的内部状况散布和灵敏的非线性过渡函数，在对高度结构化的序列数据建模的使命中现已逾越了动态贝叶斯网络，如隐马尔可夫模型（HMMs）。这些数据，例如来自语音和笔迹的数据，一般包含潜在改动要素与观测数据之间的杂乱联系。规范RNN模型在其结构中具有十分有限的随机性或可变性，来自输出条件概率模型。本文将介绍在RNN中运用高档潜在随机变量对序列数据的可变性建模的不同办法，以及在VAE（变分主动编码器）原理下这种RNN模型的练习办法。咱们将探究运用对立性办法练习变分RNN模型的或许办法。与竞赛办法相反，咱们的办法在模型练习中具有理论最优性，并供给更好的模型练习稳定性。咱们的办法还经过一个独自的对立性练习进程改进了变分推理网络中的后验近似。对TIMIT语音数据的仿真成果标明，重建丢失和依据下限收敛到同一水平，对立性练习丢失收敛到0。
摘要
【5】 Enhancing Knowledge Tracing via Adversarial Training
标题：经过对立性练习加强常识追寻
链接：https://arxiv.org/abs/2108.04430
安排：Wangxuan Institute of Computer, Technology, Peking University, Beijing, China
补白：Accepted by ACM MM 2021
摘要：咱们研讨常识追寻（KT）问题，方针是追寻学生的常识掌握情况，以便对他们未来的体现做出猜测。因为深度神经网络（DNN）具有杰出的标明才能，近年来关于KT的研讨越来越集中于探究DNN以进步KT的功能。但是，咱们的经历标明，依据DNNs的KT模型或许存在过度拟合的危险，尤其是在小数据集上，然后导致有限的泛化。在本文中，咱们运用对立练习（AT）的最新进展，提出了一种有用的依据AT的KT办法（ATKT），以增强KT模型的泛化才能，然后进步KT的极限。详细来说，咱们首要结构对立性搅扰，并将其添加到原始交互嵌入中作为对立性示例。进一步运用原始示例和对立性示例联合练习KT模型，使其不只对对立性示例具有鲁棒性，而且增强了对原始示例的泛化才能。为了更好地完结AT，咱们提出了一个高效、专心的LSTM模型作为KT骨干，其间，要害是一个主张的常识躲藏状况留意模块，该模块自习惯地调集来自从前常识躲藏状况的信息，一起杰出当时常识躲藏状况的重要性，以做出更精确的猜测。在四个公共基准数据集进步行的很多试验标明，咱们的ATKT完结了最新的功能。代码坐落：\color{blue}{\url{https://github.com/xiaopengguo/ATKT}}.
摘要：We study the problem of knowledge tracing (KT) where the goal is to trace the
students' knowledge mastery over time so as to make predictions on their future
performance. Owing to the good representation capacity of deep neural networks
(DNNs), recent advances on KT have increasingly concentrated on exploring DNNs
to improve the performance of KT. However, we empirically reveal that the DNNs
based KT models may run the risk of overfitting, especially on small datasets,
leading to limited generalization. In this paper, by leveraging the current
advances in adversarial training (AT), we propose an efficient AT based KT
method (ATKT) to enhance KT model's generalization and thus push the limit of
KT. Specifically, we first construct adversarial perturbations and add them on
the original interaction embeddings as adversarial examples. The original and
adversarial examples are further used to jointly train the KT model, forcing it
is not only to be robust to the adversarial examples, but also to enhance the
generalization over the original ones. To better implement AT, we then present
an efficient attentive-LSTM model as KT backbone, where the key is a proposed
knowledge hidden state attention module that adaptively aggregates information
from previous knowledge hidden states while simultaneously highlighting the
importance of current knowledge hidden state to make a more accurate
prediction. Extensive experiments on four public benchmark datasets demonstrate
that our ATKT achieves new state-of-the-art performance. Code is available at:
\color{blue} {\url{https://github.com/xiaopengguo/ATKT}}.
【6】 On Procedural Adversarial Noise Attack And Defense
标题：论程序性对立性噪声攻防
链接：https://arxiv.org/abs/2108.04409
安排：Received: date Accepted: date
摘要：深度神经网络（DNN）简略遭到仇视示例的进犯，这些示例会诱使神经网络在输入图画上发生小扰动时发生猜测差错。研讨人员一向致力于促进对无梯度且对数据散布知之甚少的遍及对立性扰动（UAP）的研讨。程序对立性噪声at-tack是一种无数据的通用扰动生成办法。在本文中，咱们提出了两种依据进程噪声函数的通用对立性扰动（UAP）生成办法：单纯形噪声和Worley噪声。在咱们的结构中，运用烘托技能生成搅扰视觉分类的暗影。在不改动语义标明的情况下，经过咱们的办法生成的对立性示例显现出优异的进犯功能。
摘要：Deep Neural Networks (DNNs) are vulnerable to adversarial examples which
would inveigle neural networks to make prediction errors with small per-
turbations on the input images. Researchers have been devoted to promoting the
research on the universal adversarial perturbations (UAPs) which are
gradient-free and have little prior knowledge on data distributions. Procedural
adversarial noise at- tack is a data-free universal perturbation generation
method. In this paper, we propose two universal adversarial perturbation (UAP)
generation methods based on procedural noise functions: Simplex noise and
Worley noise. In our framework, the shading which disturbs visual
classification is generated with rendering technology. Without changing the
semantic representations, the adversarial examples generated via our methods
show superior performance on the attack.
【7】 Adversarial Open Domain Adaption Framework (AODA): Sketch-to-Photo Synthesis
标题：对立性敞开范畴习惯结构(Aoda)：草图到相片组成
链接：https://arxiv.org/abs/2108.04351
安排：Department of Computer Engineering, University of Mumbai, Mumbai, MH, India
补白：8 pages, 9 figures
摘要：本文旨在证明对立性敞开域自习惯结构在草图到相片组成中的有用性。从手绘草图生成实在相片的无监督敞开域自习惯具有应战性，因为该类草图没有用于练习数据的草图。短少学习监督以及徒手画和图画范畴之间的巨大范畴距离使得学习变得困难。咱们提出了一种办法，学习从草图到相片和从相片到草图的生成，以从相片组成短少的手绘图纸。因为组成草图和实在草图之间的范畴距离，在处理短少类别的图纸时，经过假图纸练习的生成器或许会发生不令人满意的成果。为了处理这个问题，咱们供给了一种简略但有用的敞开域采样和优化办法，使生成器将虚伪图形视为实在图形。咱们的办法将所学的草图到相片和相片到草图的映射从域内输入推行到敞开域类别。在Scribble和SketchyCOCO数据集上，咱们将咱们的技能与最新的竞赛办法进行了比较。关于许多类型的敞开范畴图纸，咱们的模型在组成精确的色彩、物质和保存结构布局方面优于令人形象深入的成果。
摘要
【8】 Explainable AI and susceptibility to adversarial attacks: a case study in classification of breast ultrasound images
标题：可解说性人工智能与歹意进犯易理性：乳腺超声图画分类的事例研讨
链接：https://arxiv.org/abs/2108.04345
安排：Electrical and Computer Engineering, Concordia University, Montreal, Canada
补白：4 pages, 4 figures, Accepted to IEEE IUS 2021
摘要：超声是一种非侵入性成像办法，能够方便地用于分类可疑的乳腺结节，并有或许检测乳腺癌的发病。最近，卷积神经网络（CNN）技能在将乳腺超声图画分为良性和恶性方面显现了杰出的成果。但是，CNN推理作为一个黑箱模型，因而，其决议方案是不行解说的。因而，越来越多的人致力于解说这一进程，尤其是经过GRAD-CAM和其他技能，为CNN的内部作业供给视觉解说。除了解说外，这些办法还供给了临床上重要的信息，如确认活检或医治的方位。在这项作业中，咱们剖析了怎么规划实践上无法检测到的对立性进犯来明显改动这些重要性地图。此外，咱们将展现重要性映射的这种改动或许随同或不随同分类成果的改动，使得它们更难被检测。因而，在运用这些重要性图来说明深度学习的内部作业原理时，有必要当心。最终，咱们运用多使命学习（MTL），提出了一种依据ResNet-50的新网络，以进步分类精度。咱们的灵敏性和特异性与最先进的成果适当。
摘要：Ultrasound is a non-invasive imaging modality that can be conveniently used
to classify suspicious breast nodules and potentially detect the onset of
breast cancer. Recently, Convolutional Neural Networks (CNN) techniques have
shown promising results in classifying ultrasound images of the breast into
benign or malignant. However, CNN inference acts as a black-box model, and as
such, its decision-making is not interpretable. Therefore, increasing effort
has been dedicated to explaining this process, most notably through GRAD-CAM
and other techniques that provide visual explanations into inner workings of
CNNs. In addition to interpretation, these methods provide clinically important
information, such as identifying the location for biopsy or treatment. In this
work, we analyze how adversarial assaults that are practically undetectable may
be devised to alter these importance maps dramatically. Furthermore, we will
show that this change in the importance maps can come with or without altering
the classification result, rendering them even harder to detect. As such, care
must be taken when using these importance maps to shed light on the inner
workings of deep learning. Finally, we utilize Multi-Task Learning (MTL) and
propose a new network based on ResNet-50 to improve the classification
accuracies. Our sensitivity and specificity is comparable to the state of the
art results.
【9】 TDLS: A Top-Down Layer Searching Algorithm for Generating Counterfactual Visual Explanation
标题：TDLS：一种自顶向下生成反实际视觉解说的层查找算法
链接：https://arxiv.org/abs/2108.04238
安排：Huawei Technologies Co., Ltd, Hong Kong SAR, CHINA
补白：1 page
摘要：人工智能的解说以及算法决议方案的公平性和决议方案模型的透明度变得越来越重要。在翻开黑盒模型时，规划有用且人性化的技能至关重要。反实际契合人类的思想办法，并供给了人性化的解说，其相应的解说算法是指对给定数据点进行战略性改动，以使其模型输出“反实际”，即康复猜测。在本文中，咱们选用反实际解说的细粒度图画分类问题。咱们展现了一种自习惯办法，该办法能够经过运用自顶向下的层查找算法（TDLS）显现组成的反实际特征图来给出反实际解说。咱们现已证明，咱们的TDLS算法能够在加州理工大学UCSD Birds 200数据集上运用VGG-16模型以有用的办法供给更灵敏的反实际视觉解说。最终，咱们评论了反实际视觉解说的几种适用场景。
摘要
半/弱/无/有监督|不确认性|主动学习(3篇)
【1】 R4Dyn: Exploring Radar for Self-Supervised Monocular Depth Estimation of Dynamic Scenes
标题：R4Dyn：用于动态场景自监督单目深度估量的探究式雷达
链接：https://arxiv.org/abs/2108.04814
安排： Computer Aided Medical Procedures, Technical University of Munich (TUM), Germany, BMW Group, Munich, Germany, Computer Aided Medical Procedures, Johns Hopkins University (JHU), Baltimore, MD, USA, Google, Z¨urich, Switzerland
补白：Currently under review
摘要：尽管驾驭场景中的自我监督单目深度估量与监督办法的功能适当，但违背静态世界假定仍或许导致交通参加者的过错深度猜测，然后构成潜在的安全问题。在本文中，咱们介绍了R4Dyn，这是一种在自监督深度估量结构上运用低本钱雷达数据的新技能。特别是，咱们展现了怎么在练习期间将雷达用作弱监督信号，以及怎么作为额定的输入来增强推理时的估量鲁棒性。因为轿车雷达随时可用，因而能够从各种现有车辆搜集练习数据。此外，经过滤波和扩展信号，使其与依据学习的办法兼容，咱们处理了雷达固有的问题，如噪声和稀少性。运用R4Dyn，咱们能够战胜自监督深度估量的一个首要约束，即交通参加者的猜测。在具有应战性的nuScenes数据集上，咱们大幅进步了对动态方针（如轿车）的估量37%，然后证明晰雷达是用于自主车辆单目深度估量的有价值的附加传感器。此外，咱们方案揭露该代码。
摘要
【2】 Semi-supervised classification of radiology images with NoTeacher: A Teacher that is not Mean
标题：用NoTeacher完结放射学图画的半监督分类：一位不尖刻的教师
链接：https://arxiv.org/abs/2108.04423
安排：Krishnaswamya,∗, Institute for Infocomm Research, Agency for Science Technology and Research (ASTAR), Singapore, National University of Singapore, Singapore
补白：Preprint submitted to Medical Image Analysis. Accepted in June 2021
摘要：深度学习模型在放射图画分类中取得了很好的功能，但因为需求很多的符号练习数据集，其实践运用遭到了约束。半监督学习（SSL）办法运用较小的符号数据集和较大的未符号数据集，并供给了下降符号本钱的潜力。在这项作业中，咱们介绍了NoTeacher，一种新的依据一致性的SSL结构，它结合了概率图形模型。与Mean Teacher不同的是，Mean Teacher经过时刻调集保护更新的教师网络，NoTeacher选用了两个独立的网络，然后消除了对教师网络的需求。咱们将演示怎么定制NoTeacher，以应对放射学图画分类中的一系列应战。详细来说，咱们描绘了对2D和3D输入、单标签和多标签分类以及练习数据中有标签和无标签部分之间的类别散布不匹配的场景的习惯。在对三个公共基准数据集进行的实际实证点评中，涵盖了放射学的首要形式（X射线、CT、MRI），咱们标明NoTeacher在不到5-15%的符号预算下完结了90-95%的彻底监督AUROC。此外，NoTeacher以最小的超参数调整优于已树立的SSL办法，而且作为放射学运用中的半监督学习的一个原则性和有用的选项具有含义。
摘要
【3】 Active Learning for Transition State Calculation
标题：用于过渡态核算的主动学习
链接：https://arxiv.org/abs/2108.04698
安排：a College of Big Data and Internet, Shenzhen Technology University, Shenzhen , People’s Republic of China, School of Mathematics and Statistics, Central South University, Changsha , People’s Republic of China, School of Data Science and Department of Mathematics
补白：27 pages
摘要：过渡态（TS）核算是核算密集型能量函数的一大应战。传统的办法需求在很多的方位核算能量函数的梯度。为了削减实在梯度的贵重核算量，咱们提出了一个主动学习结构，该结构由计算代替模型、能量函数的高斯进程回归（GPR）和马鞍型过渡态的单walker动力学办法、柔软重音动力学（GAD）组成。TS由运用于梯度向量和Hessian矩阵的GPR署理的GAD检测。咱们进步功率的要害要素是一种主动学习办法，该办法按次序规划信息量最大的方位，并在这些方位对原始模型进行点评，以练习探地雷达。咱们将这种主动学习使命描绘为最优试验规划问题，并提出了一种依据样本的次优原则来结构最优方位。咱们标明，新办法明显削减了原始模型所需的能量或力点评数量。
摘要
搬迁|Zero/Few/One-Shot|自习惯(2篇)
【1】 AdaRNN: Adaptive Learning and Forecasting of Time Series
标题：AdaRNN：时刻序列的自习惯学习与猜测
链接：https://arxiv.org/abs/2108.04443
安排：Nanjing University, Nanjing, China, Microsoft Research Asia, Beijing, China, Institute of Data Science, National University of Singapore, Nanyang Technological University, Zhejiang University
补白：Accepted by CIKM 2021 as a full paper; 10 pages; code at: this https URL
摘要
摘要
【2】 Deep Transfer Learning for Identifications of Slope Surface Cracks
标题：深度搬迁学习在坡面裂缝辨认中的运用
链接：https://arxiv.org/abs/2108.04235
安排：School of Engineering and Technology, China University of Geosciences (Beijing), China
摘要：滑坡等地质灾害给公民生命产业安全构成了巨大丢失，常伴有地表裂缝。假如能够及时辨认这些地表裂缝，对地质灾害的监测和预警具有重要含义。现在，最常用的裂纹辨认办法是人工检测，功率和精确率都很低。为了对滑坡等地质灾害进行快速监测和预警，本文提出了一种能够有用辨认边坡外表裂缝的深度搬迁学习结构。其根本思想是经过练习（a）混凝土裂缝的大样本数据集和（b）土壤和岩体裂缝的小样本数据集来选用搬迁学习。在该结构中，（1）依据混凝土裂缝的大样本数据集构建了预练习裂缝辨认模型(2） 在此根底上，进一步构建了依据岩土体裂缝小样本数据集的精密裂缝辨认模型。主张的结构可用于在高陡坡进步行无人机勘察，以完结滑坡监测和预警，保证公民生命和产业安全。
摘要
强化学习(5篇)
【1】 A Survey on Deep Reinforcement Learning for Data Processing and Analytics
标题：用于数据处理和剖析的深度强化学习总述
链接：https://arxiv.org/abs/2108.04526
安排：†National University of Singapore, § Zhejiang University, ‡Beijing Institute of Techonology
补白：39 pages, 5 figures and 2 tables
摘要：数据处理和剖析是根底和遍及的。算法在数据处理和剖析中起着至关重要的作用，许多算法规划结合了人类常识和经历的启示式和一般规矩，以进步其有用性。近年来，强化学习，特别是深度强化学习（DRL）在许多范畴得到了越来越多的探究和开发，因为它能够在杂乱的环境中学习比静态规划的算法更好的战略。受这一趋势的推进，咱们全面回忆了最近的作业，要点是运用深度强化学习来改进数据处理和剖析。首要，咱们介绍了深层强化学习的要害概念、理论和办法。接下来，咱们将评论在数据库体系上的深度强化学习布置，在各个方面促进数据处理和剖析，包含数据安排、调度、调优和索引。然后，咱们查询了深度强化学习在数据处理和剖析中的运用，从数据预备、天然语言界面到医疗保健、金融科技等。最终，咱们评论了在数据处理和剖析中运用深度强化学习的重要敞开应战和未来研讨方向。
摘要
【2】 Deep Reinforcement Learning for Demand Driven Services in Logistics and Transportation Systems: A Survey
标题：物流运输体系需求驱动服务的深度强化学习研讨总述
链接：https://arxiv.org/abs/2108.04462
安排： Li are with Beijing NationalResearch Center for Information Science and Technology (BNRist) andwith Department of Electronic Engineering, Tsinghua University
补白：21 pages. survey preprint
摘要：最近的技能开展为城市生活带来了很多新的需求驱动服务（DDS），包含搭车同享、按需配送、快递体系和仓储。在DDS中，服务环路是一个根本结构，包含其服务工、服务供给者和相应的服务方针。服务人员应将人员或包裹从供货商运送至方针方位。因而，DDS中的各种规划使命可分为两个独自的阶段：1）调度，即依据需求/供给散布构成服务环路；2）路由，即在构建的环路内确认特定的服务订单。在这两个阶段生成高质量的战略关于开发DDS很重要，但面对几个应战。一起，深度强化学习（DRL）近年来得到了迅速开展。它是处理这些问题的有力东西，因为DRL能够学习参数化模型，而无需依靠太多依据问题的假定，并经过学习次序决议方案优化长时刻作用。在本总述中，咱们首要界说了DDS，然后要点介绍了DDS中的常见运用和重要决议方案/操控问题。针对每个问题，咱们全面介绍了现有的DRL处理方案，并在\text中进一步总结{https://github.com/tsinghua-fib-lab/DDS\_查询}。咱们还介绍了开发和点评DDS运用程序的敞开仿真环境。最终，咱们剖析了依然存在的应战，并评论了DDS DRL处理方案的进一步研讨时机。
摘要
【3】 High Quality Related Search Query Suggestions using Deep Reinforcement Learning
标题：依据深度强化学习的高质量相关查找查询主张
链接：https://arxiv.org/abs/2108.04452
安排：LinkedIn, Sunnyvale, California, USA
补白：Multi-Armed Bandits and Reinforcement Learning: Advancing Decision Making in E-Commerce and Beyond at KDD 2021
摘要：“高质量相关查找查询主张”使命旨在引荐实在、精确、多样、相关和吸引人的查找查询。获取很多查询质量的人工注释是贵重的。从前关于有监督的查询主张模型的作业存在挑选和露出差错，而且依靠于稀少和喧闹的即时用户反应（例如，点击），导致低质量的主张。强化学习技能用于运用查找成果中的术语从头结构查询，其可扩展性有限，无法运用于大规模职业运用。为了引荐高质量的相关查找查询，咱们练习了一个深度强化学习模型来猜测用户下一步将输入的查询。奖赏信号由依据长时刻会话的用户反应、语法关联性和生成查询的估量天然度组成。与基线监督模型比较，咱们提出的办法在引荐多样性（3%）、下流用户参加度（4.2%）和每句话单词重复率（82%）方面完结了明显的相对改进。
摘要
【4】 Adaptable image quality assessment using meta-reinforcement learning of task amenability
标题：依据使命习惯性元强化学习的自习惯图画质量点评
链接：https://arxiv.org/abs/2108.04359
安排： Centre for Medical Image Computing, WellcomeEPSRC Centre for Interventional, Surgical Sciences, and Department of Medical Physics Biomedical Engineering, University College London, London, UK, InstaDeep, London, UK
补白：Accepted at ASMUS 2021 (The 2nd International Workshop of Advances in Simplifying Medical UltraSound)
摘要：许多医学图画剖析使命的功能与图画数据质量密切相关。在开发现代深度学习算法时，不依靠于片面（依据人的）图画质量点评（IQA），使命习惯性或许供给使命特定图画质量的客观衡量。为了猜测使命习惯性，IQA署理运用强化学习（RL）和一起优化的使命猜测器（如分类或分段神经网络）进行练习。在这项作业中，咱们开发了搬迁学习或习惯战略，以进步IQA署理和使命猜测的习惯性，然后削减对高质量、专家符号的练习数据的依靠。所提出的搬迁学习战略在元强化学习（meta-RL）结构中从头表述了使命习惯性的原始RL问题。由此发生的算法有助于agent有用地习惯不同的图画质量界说，每个界说都有自己的马尔可夫决议方案进程环境，包含不同的图画、标签和自习惯使命猜测器。咱们的作业标明，在非专家使命标签上预先练习的IQA署理能够依据专家使命标签的界说，仅运用一小部分专家标签来猜测使命的习惯性。运用249名前列腺癌患者的6644张临床超声图画，咱们对图画分类和切割使命的成果标明，所提出的IQA办法能够运用数据进行调整，别离只要19.7%和29.6%的专家评定一致性标签，依然能够完结可比的IQA和使命功能，不然，需求一个100%专家标签的练习数据集。
摘要：The performance of many medical image analysis tasks are strongly associated
with image data quality. When developing modern deep learning algorithms,
rather than relying on subjective (human-based) image quality assessment (IQA),
task amenability potentially provides an objective measure of task-specific
image quality. To predict task amenability, an IQA agent is trained using
reinforcement learning (RL) with a simultaneously optimised task predictor,
such as a classification or segmentation neural network. In this work, we
develop transfer learning or adaptation strategies to increase the adaptability
of both the IQA agent and the task predictor so that they are less dependent on
high-quality, expert-labelled training data. The proposed transfer learning
strategy re-formulates the original RL problem for task amenability in a
meta-reinforcement learning (meta-RL) framework. The resulting algorithm
facilitates efficient adaptation of the agent to different definitions of image
quality, each with its own Markov decision process environment including
different images, labels and an adaptable task predictor. Our work demonstrates
that the IQA agents pre-trained on non-expert task labels can be adapted to
predict task amenability as defined by expert task labels, using only a small
set of expert labels. Using 6644 clinical ultrasound images from 249 prostate
cancer patients, our results for image classification and segmentation tasks
show that the proposed IQA method can be adapted using data with as few as
respective 19.7% and 29.6% expert-reviewed consensus labels and still achieve
comparable IQA and task performance, which would otherwise require a training
dataset with 100% expert labels.
【5】 Imitation Learning by Reinforcement Learning
标题：强化学习中的仿照学习
链接：https://arxiv.org/abs/2108.04763
摘要：仿照学习算法从专家行为的演示中学习战略。咱们发现，关于确认性专家来说，仿照学习能够经过简化为强化学习来完结，而强化学习一般被以为更为困难。咱们进行的试验证明，咱们的还原在接连操控使命的实践中作用杰出。
摘要：Imitation Learning algorithms learn a policy from demonstrations of expert
behavior. Somewhat counterintuitively, we show that, for deterministic experts,
imitation learning can be done by reduction to reinforcement learning, which is
commonly considered more difficult. We conduct experiments which confirm that
our reduction works well in practice for a continuous control task.
元学习(1篇)
【1】 Meta-repository of screening mammography classifiers
标题：挑选乳房X线拍摄分类器的元库
链接：https://arxiv.org/abs/2108.04800
安排：Center for Data Science, New York University, NY, USA, Department of Radiology, NYU Langone Health, NY, USA, Center for Advanced Imaging Innovation and Research, NYU Langone Health, NY, USA, Jagiellonian, University, Kraków, Poland
补白：16 pages, 2 figures. Meta-repository available at this https URL
摘要：人工智能（AI）正在改动医学，并有望改进临床确诊。在乳腺癌筛查方面，最近的几项研讨标明人工智能有或许进步放射科医师的精确性，然后有助于前期癌症确诊和削减不必要的查看。跟着提出的模型数量及其杂乱性的添加，为了重现成果和比较不同的办法，从头完结它们变得越来越困难。为了使这一运用范畴的研讨具有可重复性，并使不同办法之间能够进行比较，咱们发布了一个元存储库，其间包含用于筛查乳房X光片分类的深度学习模型。这个元存储库创建了一个结构，能够对任何私家或公共筛查乳腺X光拍摄数据集的机器学习模型进行点评。从一开端，咱们的元存储库就包含五个最先进的模型，具有开源完结和跨渠道兼容性。咱们比较他们的体现在五个世界数据集：两个私家纽约大学乳腺癌筛查数据集，以及三个公共（DDSM，乳房和我国乳房X光拍摄数据库）数据集。咱们的结构具有灵敏的规划，能够推行到其他医学图画剖析使命。元存储库坐落https://www.github.com/nyukat/mammography_metarepository.
摘要
分层学习(1篇)
【1】 Hierarchical Latent Relation Modeling for Collaborative Metric Learning
标题：面向协作衡量学习的层次潜在联系建模
链接：https://arxiv.org/abs/2108.04655
补白：15th ACM Conference on Recommender Systems (RecSys 2021)
摘要：协作衡量学习（CML）是近年来鼓起的一种依据内隐反应协同过滤的引荐形式。但是，规范的CML办法学习固定的用户和项方针明，这无法捕获用户的杂乱爱好。CML的现有扩展也疏忽了用户项目联系的异质性，即用户能够一起喜爱十分不同的项目，或许疏忽了潜在的项目联系，即用户对项意图偏好不只取决于其内涵特征，还取决于他们曾经交互过的项目。在本文中，咱们提出了一个分层CML模型，该模型从隐式数据中联合捕获潜在的用户项和项项联系。咱们的办法遭到常识图嵌入的翻译机制的启示，并运用依据回忆的留意网络。咱们经过在多个实在数据集的引荐使命上优于现有的CML模型，从经历上证明晰这种联合联系模型的相关性。咱们的试验还着重了当时CML联系模型在十分稀少的数据集上的局限性。
摘要
医学相关(5篇)
【1】 Analyzing Effects of The COVID-19 Pandemic on Road Traffic Safety: The Cases of New York City, Los Angeles, and Boston
标题：剖析冠状病毒大盛行对路途交通安全的影响--以纽约市、洛杉矶和波士顿为例
链接：https://arxiv.org/abs/2108.04787
安排：Department of Computer Science, University of Memphis
摘要：新冠病毒-19大盛行在全世界构成了严重的社会和经济影响。除了对健康的影响外，对交通行为的影响也是突但是剧烈的。咱们剖析了流感大盛行以及相应的地方政府指令和约束对美国纽约市、洛杉矶和波士顿路途交通安全的影响。详细而言，咱们研讨了大盛行迸发前后事端热门的散布，发现与从前比较，交通事端在地址和时刻上都发生了改动。此外，咱们还研讨了这些热门区域的路途网络特征，期望了解热门搬运的根本原因。
摘要：The COVID-19 pandemic has resulted in significant social and economic impacts
throughout the world. In addition to the health consequences, the impacts on
traffic behaviors have also been sudden and dramatic. We have analyzed how the
road traffic safety of New York City, Los Angeles, and Boston in the U.S. have
been impacted by the pandemic and corresponding local government orders and
restrictions. To be specific, we have studied the accident hotspots'
distributions before and after the outbreak of the pandemic and found that
traffic accidents have shifted in both location and time compared to previous
years. In addition, we have studied the road network characteristics in those
hotspot regions with the hope to understand the underlying cause of the hotspot
shifts.
【2】 U-Net-and-a-half: Convolutional network for biomedical image segmentation using multiple expert-driven annotations
标题：U-Net-1.5：运用多专家驱动注释的卷积网络用于生物医学图画切割
链接：https://arxiv.org/abs/2108.04658
安排：Section of Computational Biomedicine, Department of Medicine, Boston University School of Medicine, Department of Pathology, Academic Medical Center, Meibergdreef , AZ Amsterdam, The
补白：this https URL
摘要：开发用于生物医学切割的深度学习系一致般需求拜访专家驱动、手动注释的数据集。假如同一幅图画的注释触及多个专家，那么专家间的一致性并不一定完美，而且没有一个专家注释能够精确地捕捉一切图画上感爱好区域的所谓根本实际。此外，运用来自多个专家的注释生成参阅估量值也不是一件小事。在这里，咱们提出了一个深度神经网络，界说为U-Net-a-半月，它能够一起从多个专家对同一组图画履行的注释中学习。U-Net-a-半月包含用于从输入图画生成特征的卷积编码器、答应从由多个专家独立生成的注释中取得的图画掩码一起学习的多个解码器，以及同享的低维特征空间。为了证明咱们的结构的适用性，咱们别离运用了来自数字病理学和放射学的两个不同的数据集。详细而言，咱们别离运用病理学家驱动的人类肾脏活检全片图画上的肾小球注释（10例患者）和放射科医师驱动的血管内超声图画上的人类动静脉瘘管腔横截面注释（10例患者）来练习两个独立的模型。依据U-Net和半U-Net的模型超过了仅依据单个专家注释练习的传统U-Net模型的功能，然后扩展了生物医学图画切割中多使命学习的规模。
摘要
【3】 Known Operator Learning and Hybrid Machine Learning in Medical Imaging --- A Review of the Past, the Present, and the Future
标题：医学印象学中的已知算子学习和混合机器学习-曩昔、现在和未来的回忆
链接：https://arxiv.org/abs/2108.04543
安排：Friedrich-Alexander-University Erlangen-Nuremberg, Germany
补白：22 pages, 4 figures, submitted to "Progress in Biomedical Engineering"
摘要
摘要
【4】 A Survey of Machine Learning Techniques for Detecting and Diagnosing COVID-19 from Imaging
标题：从印象中检测和确诊冠状病毒的机器学习技能总述
链接：https://arxiv.org/abs/2108.04344
安排：Department of Computer Science Engineering, Stamford University, Dhaka , Bangladesh, School of Computing and Mathematics, Charles Sturt University, NSW , Australia, Department of Computer Science Engineering, University of Chittagong, Chittagong
补白：None
摘要：因为逆转录聚合酶链反应（RT-PCR）检测的可用性有限且本钱高，许多研讨提出了从医学印象中检测新冠病毒19的机器学习技能。本研讨的意图是体系地回忆、点评和归纳运用不同机器学习技能从胸部X射线和CT扫描图画检测和确诊新冠病毒-19的研讨文章。在相关书目数据库中进行了结构化的文献检索，以保证查询仅以可重复和高质量的研讨为中心。咱们依据当选规范挑选论文。在本次查询中，咱们检查了契合归入规范的98美元的文章。咱们查询了一整套与新冠病毒相关的胸部印象剖析技能，包含数据搜集、预处理、特征提取、分类和可视化。咱们以为CT扫描和X射线都被广泛用于描绘检测新冠病毒19的医学成像的最新开展。这项查询为研讨人员供给了关于不同机器学习技能及其在胸部成像检测和确诊新冠病毒19方面的体现的有价值的见地。最终，评论了运用机器学习技能检测新冠病毒的应战和局限性以及未来的研讨方向。
摘要：Due to the limited availability and high cost of the reverse
transcription-polymerase chain reaction (RT-PCR) test, many studies have
proposed machine learning techniques for detecting COVID-19 from medical
imaging. The purpose of this study is to systematically review, assess, and
synthesize research articles that have used different machine learning
techniques to detect and diagnose COVID-19 from chest X-ray and CT scan images.
A structured literature search was conducted in the relevant bibliographic
databases to ensure that the survey solely centered on reproducible and
high-quality research. We selected papers based on our inclusion criteria. In
this survey, we reviewed $98$ articles that fulfilled our inclusion criteria.
We have surveyed a complete pipeline of chest imaging analysis techniques
related to COVID-19, including data collection, pre-processing, feature
extraction, classification, and visualization. We have considered CT scans and
X-rays as both are widely used to describe the latest developments in medical
imaging to detect COVID-19. This survey provides researchers with valuable
insights into different machine learning techniques and their performance in
the detection and diagnosis of COVID-19 from chest imaging. At the end, the
challenges and limitations in detecting COVID-19 using machine learning
techniques and the future direction of research are discussed.
【5】 Automated Olfactory Bulb Segmentation on High Resolutional T2-Weighted MRI
标题：依据高分辨率T2加权MRI的嗅球主动切割
链接：https://arxiv.org/abs/2108.04267
