机器学习学术速递[2021.9.16]

机器学习学术速递[2021.9.16]

访问arxivdaily.com获取含摘要速递,更有收藏、搜索等功能,涵盖CS|物理|数学|经济|统计|金融|生物|电气领域
同步公众号:arXiv每日学术速递,欢迎关注

cs.LG 方向,今日共计75篇


Graph相关(图学习|图神经网络|图优化等)(3篇)

【1】 Fusion with Hierarchical Graphs for Mulitmodal Emotion Recognition
基于层次图融合的多模态情感识别
链接arxiv.org/abs/2109.0714
作者:Shuyun Tang,Zhaojie Luo,Guoshun Nan,Yuichiro Yoshikawa,Ishiguro Hiroshi
机构: UC Berkeley , Osaka University , Singapore University of Technology and Design

【2】 Graph Embedding via Diffusion-Wavelets-Based Node Feature Distribution Characterization
基于扩散小波节点特征分布特征的图嵌入
链接arxiv.org/abs/2109.0701
作者:Lili Wang,Chenghan Huang,Weicheng Ma,Xinyuan Cao,Soroush Vosoughi
机构:Dartmouth College, Hanover, New Hampshire, USA, Millennium Management, LLC, New York, New York, USA, Georgia Institute of Technology, Atlanta, Georgia, USA
备注:In CIKM 2021

【3】 HeMI: Multi-view Embedding in Heterogeneous Graphs
HEMI:异构图中的多视图嵌入
链接arxiv.org/abs/2109.0700
作者:Costas Mavromatis,George Karypis

Transformer(2篇)

【1】 Matching with Transformers in MELT
与熔融Transformer的匹配
链接arxiv.org/abs/2109.0740
作者:Sven Hertling,Jan Portisch,Heiko Paulheim
机构: Data and Web Science Group, University of Mannheim, Germany, SAP SE Business Technology Platform - One Domain Model, Walldorf, Germany
备注:accepted at the Ontology Matching Workshop at the International Semantic Web Conference (ISWC 2021)

【2】 Explainable Identification of Dementia from Transcripts using Transformer Networks
利用Transformer网络从成绩单中解释痴呆的识别
链接arxiv.org/abs/2109.0698
作者:Loukas Ilias,Dimitris Askounis
机构:School of Electrical and Computer Engineering, National Technical University of Athens, Zografou, Athens
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

GAN|对抗|攻击|生成相关(4篇)

【1】 NBcoded: network attack classifiers based on Encoder and Naive Bayes model for resource limited devices
NBCoded:基于编码器和朴素贝叶斯模型的资源受限设备网络攻击分类器
链接arxiv.org/abs/2109.0727
作者:Lander Segurola-Gil,Francesco Zola,Xabier Echeberria-Barrio,Raul Orduna-Urrutia
机构:Orduna-Urrutia, Vicomtech Foundation, Basque Research and Technology Alliance (BRTA), Institute of Smart Cities, Public University of Navarre, Pamplona, Spain
备注:It will be published in "Communications in Computer and Information Science" and presented in the 3rd Workshop of Machine Learning for Cybersecurity (MLCS)

【2】 Adversarial Mixing Policy for Relaxing Locally Linear Constraints in Mixup
局部松弛混合中线性约束的对抗性混合策略
链接arxiv.org/abs/2109.0717
作者:Guang Liu,Yuzhao Mao,Hailong Huang,Weiguo Gao,Xuan Li
机构:PingAn Life Insurance of China
备注:This paper is accepted to appear in the main conference of EMNLP2021

【3】 Balancing detectability and performance of attacks on the control channel of Markov Decision Processes
平衡马尔可夫决策过程控制信道上攻击的可检测性和性能
链接arxiv.org/abs/2109.0717
作者:Alessio Russo,Alexandre Proutiere
机构:Division of Decision and Control Systems, EECS School, KTH Royal Institute of Technology, Stockholm

【4】 Universal Adversarial Attack on Deep Learning Based Prognostics
基于深度学习的预测学的通用对抗性攻击
链接arxiv.org/abs/2109.0714
作者:Arghya Basak,Pradeep Rathore,Sri Harsha Nistala,Sagar Srinivas,Venkataramana Runkana
机构:TCS Research, Pune, India
备注:7 pages

半/弱/无/有监督|不确定性|主动学习(4篇)

【1】 SupCL-Seq: Supervised Contrastive Learning for Downstream Optimized Sequence Representations
SupCL-Seq:下游优化序列表示的有监督对比学习
链接arxiv.org/abs/2109.0742
作者:Hooman Sedghamiz,Shivam Raval,Enrico Santus,Tuka Alhanai,Mohammad Ghassemi
机构: DSIG - Bayer Pharmaceuticals, New Jersey, USA, New York University, Abu Dhabi, UAE, Michigan State University, Michigan, USA
备注:short paper, EMNLP 2021, Findings

【2】 Improving Robustness and Efficiency in Active Learning with Contrastive Loss
利用对比损失提高主动学习的鲁棒性和效率
链接arxiv.org/abs/2109.0687
作者:Ranganath Krishnan,Nilesh Ahuja,Alok Sinha,Mahesh Subedar,Omesh Tickoo,Ravi Iyer
机构: Intel Labs, Hillsboro, USA, † Intel Corporation, Bangalore, India
备注:arXiv admin note: substantial text overlap with arXiv:2109.06321

【3】 The potential of self-supervised networks for random noise suppression in seismic data
自监督网络抑制地震数据随机噪声的潜力
链接arxiv.org/abs/2109.0734
作者:Claire Birnie,Matteo Ravasi,Tariq Alkhalifah,Sixiu Liu
机构:KAUST, Thuwal, Kingdom of Saudi Arabia

【4】 Evolutionary Reinforcement Learning Dynamics with Irreducible Environmental Uncertainty
具有不可约环境不确定性的进化强化学习动力学
链接arxiv.org/abs/2109.0725
作者:Wolfram Barfuss,Richard P. Mann
机构: 20 10)Preprint working paper 1University of Tübingen, Germany; 2University of Leeds
备注:14 pages, 7 figures

强化学习(4篇)

【1】 DCUR: Data Curriculum for Teaching via Samples with Reinforcement Learning
DCUR:基于强化学习的样本教学数据课程
链接arxiv.org/abs/2109.0738
作者:Daniel Seita,Abhinav Gopal,Zhao Mandi,John Canny
机构: [ 3 2] and whenthe student can engage in a small amount of self-generated 1University of California
备注:Supplementary material is available at this https URL

【2】 Back to Basics: Deep Reinforcement Learning in Traffic Signal Control
回归基础:深度强化学习在交通信号控制中的应用
链接arxiv.org/abs/2109.0718
作者:Sierk Kanis,Laurens Samson,Daan Bloembergen,Tim Bakker
机构:University of Amsterdam, Amsterdam, The Netherlands, CTO, City of Amsterdam
备注:9 pages, 4 figures; code for this paper is available at this https URL

【3】 Optimal Cycling of a Heterogenous Battery Bank via Reinforcement Learning
基于强化学习的异质电池组优化循环
链接arxiv.org/abs/2109.0713
作者:Vivek Deulkar,Jayakrishnan Nair
机构:Dept. of Electrical Engineering, IIT Bombay
备注:Appeared on IEEE SmartGridComm 2021 conference

【4】 WaveCorr: Correlation-savvy Deep Reinforcement Learning for Portfolio Management
WaveCorr:用于投资组合管理的相关性感知深度强化学习
链接arxiv.org/abs/2109.0700
作者:Saeed Marzban,Erick Delage,Jonathan Yumeng Li,Jeremie Desgagne-Bouchard,Carl Dussault
机构:GERAD & Department of Decision Sciences, HEC Montréal, Montreal, Canada, Telfer School of Management, University of Ottawa, Ottawa, Canada, Evovest, Montreal, Canada

医学相关(3篇)

【1】 Modelling Major Disease Outbreaks in the 21st Century: A Causal Approach
模拟21世纪重大疾病暴发:一种因果方法
链接arxiv.org/abs/2109.0726
作者:Abli Marathe,Saloni Parekh,Harsh Sakhrani
机构:Dept. of Information Technology, Pune Institute of Computer, Pune, India
备注:Accepted at Special Interest Group on Knowledge Discovery and Data Mining (SIGKDD-epiDAMIK) 2021: The 4th International Workshop on Epidemiology meets Data Mining and Knowledge discovery

【2】 WIP: Medical Incident Prediction Through Analysis of Electronic Medical Records Using Machine Lerning: Fall Prediction
WIP:通过机器学习分析电子病历的医疗事件预测:秋季预测
链接arxiv.org/abs/2109.0710
作者:Atsushi Yanagisawa,Chintaka Premachandra,Hiruharu Kawanaka,Atsushi Inoue,Takeo Hata,Eiichiro Ueda
机构:,Takatsuki High School, Japan, ,Shibaura Institute of Technology, Japan, ,Mie University, Japan, ,Osaka Medical and Pharmaceutical University, Japan
备注:None

【3】 Deploying clinical machine learning? Consider the following...
部署临床机器学习?请考虑以下几点。
链接arxiv.org/abs/2109.0691
作者:Charles Lu,Ken Chang,Praveer Singh,Stuart Pomerantz,Sean Doyle,Sujay Kakarmath,Christopher Bridge,Jayashree Kalpathy-Cramer
机构:Massachusetts General Hospital, Boston, MA, Massachusetts Institute of Technology, Cambridge, MA, Harvard Medical School, Boston, MA, Mass General Brigham, Boston, MA

蒸馏|知识提取(2篇)

【1】 Constraint based Knowledge Base Distillation in End-to-End Task Oriented Dialogs
端到端任务导向对话中基于约束的知识库提炼
链接arxiv.org/abs/2109.0739
作者:Dinesh Raghu,Atishya Jain,Mausam,Sachindra Joshi
机构: IIT Delhi, New Delhi, India, IBM Research, New Delhi, India
备注:D. Raghu and A. Jain contributed equally to this work

【2】 {E}fficient{BERT}: Progressively Searching Multilayer Perceptron via Warm-up Knowledge Distillation
{E}高效{BERT}:通过预热知识蒸馏逐步搜索多层感知器
链接arxiv.org/abs/2109.0722
作者:Chenhe Dong,Guangrun Wang,Hang Xu,Jiefeng Peng,Xiaozhe Ren,Xiaodan Liang
机构: Shenzhen Campus of Sun Yat-sen University , University of Oxford, Huawei Noah’s Ark Lab , DarkMatter AI Research
备注:Findings of EMNLP 2021

聚类(1篇)

【1】 Powered Hawkes-Dirichlet Process: Challenging Textual Clustering using a Flexible Temporal Prior
Powered Hawkes-Dirichlet过程:使用灵活的时间先验挑战文本聚类
链接arxiv.org/abs/2109.0717
作者:Gaël Poux-Médard,Julien Velcin,Sabine Loudcher
机构:ERIC Lab, Universit´e de Lyon, Lyon, France, -,-,-,X

自动驾驶|车辆|车道检测等(1篇)

【1】 Risk Measurement, Risk Entropy, and Autonomous Driving Risk Modeling
风险度量、风险熵与自主驾驶风险建模
链接arxiv.org/abs/2109.0721
作者:Jiamin Yu
机构:Received: date Accepted: date
备注:11 pages, 5 figures, IME 2021

联邦学习|隐私保护|加密(1篇)

【1】 Federated Learning of Molecular Properties in a Heterogeneous Setting
异质环境下分子性质的联合学习
链接arxiv.org/abs/2109.0725
作者:Wei Zhu,Andrew White,Jiebo Luo
机构:Department of Computer Science, University of Rochester, Department of Chemical Engineering, University of Rochester

推理|分析|理解|解释(3篇)

【1】 FORTAP: Using Formulae for Numerical-Reasoning-Aware Table Pretraining
FORTAP:使用公式进行数值推理的表预训练
链接arxiv.org/abs/2109.0732
作者:Zhoujun Cheng,Haoyu Dong,Fan Cheng,Ran Jia,Pengfei Wu,Shi Han,Dongmei Zhang
备注:Work in progress

【2】 Internet of Behavior (IoB) and Explainable AI Systems for Influencing IoT Behavior
行为互联网(IoB)和影响物联网行为的可解释人工智能系统
链接arxiv.org/abs/2109.0723
作者:Haya Elayan,Moayad Aloqaily,Mohsen Guizani
机构:!
备注:Submitted to IEEE Network

【3】 Non-Asymptotic Analysis of Stochastic Approximation Algorithms for Streaming Data
流数据随机逼近算法的非渐近分析
链接arxiv.org/abs/2109.0711
作者:Antoine Godichon-Baggioni,Nicklas Werge,Olivier Wintenberger
机构:LPSM, Sorbonne Université, place Jussieu, Paris, France

检测相关(2篇)

【1】 Assisting the Human Fact-Checkers: Detecting All Previously Fact-Checked Claims in a Document
协助人类事实核查人员:检测文档中所有先前经过事实核查的声明
链接arxiv.org/abs/2109.0741
作者:Shaden Shaar,Firoj Alam,Giovanni Da San Martino,Preslav Nakov
机构: Qatar Computing Research Institute, HBKU, Doha, Qatar, University of Padova, Italy
备注:detecting previously fact-checked claims, fact-checking, disinformation, fake news, social media, political debates

【2】 Photon detection probability prediction using one-dimensional generative neural network
基于一维产生式神经网络的光子探测概率预测
链接arxiv.org/abs/2109.0727
作者:Wei Mu,Alexander I. Himmel,Bryan Ramson
机构:Neutrino Division, Fermi National Accelerator Laboratory, Wilson Street and Kirk Road, Batavia, Illinois, U.S.A.

分类|识别(1篇)

【1】 Embedding Convolutions for Short Text Extreme Classification with Millions of Labels
嵌入卷积的百万标签短文本极端分类算法
链接arxiv.org/abs/2109.0731
作者:Siddhant Kharbanda,Atmadeep Banerjee,Akash Palrecha,Rohit Babbar
机构:Department of Computer Science, Aalto University

表征(2篇)

【1】 Comparing Text Representations: A Theory-Driven Approach
文本表征比较:一种理论驱动的方法
链接arxiv.org/abs/2109.0745
作者:Gregory Yauney,David Mimno
机构:Cornell University
备注:None

【2】 Deep Bregman Divergence for Contrastive Learning of Visual Representations
视觉表征对比学习的深度Bregman发散
链接arxiv.org/abs/2109.0745
作者:Mina Rezaei,Farzin Soleymani,Bernd Bischl,Shekoofeh Azizi
机构: Department of Statistics, LMU Munich, Germany, Department of Electrical and Computer Engineering, Technical University of Munich, Germany, Google Research, United States

优化|敛散性(4篇)

【1】 DROMO: Distributionally Robust Offline Model-based Policy Optimization
Dromo:基于离线模型的分布式健壮策略优化
链接arxiv.org/abs/2109.0727
作者:Ruizhen Liu,Dazhi Zhong,Zhicong Chen
机构:Affiliated High School of South China Normal University
备注:Under review of S.-T. Yau Award 2021 of Computer Science

【2】 Convergence of a Human-in-the-Loop Policy-Gradient Algorithm With Eligibility Trace Under Reward, Policy, and Advantage Feedback
奖励、政策和优势反馈下带资格追踪的人在环策略梯度算法的收敛性
链接arxiv.org/abs/2109.0705
作者:Ishaan Shah,David Halpern,Kavosh Asadi,Michael L. Littman
机构:Equal contribution 1Department of Computer Science, BrownUniversity
备注:Accepted into ICML 2021 workshops Human-AI Collaboration in Sequential Decision-Making and Human in the Loop Learning

【3】 Neural network optimal feedback control with enhanced closed loop stability
闭环稳定性增强的神经网络最优反馈控制
链接arxiv.org/abs/2109.0746
作者:Tenavi Nakamura-Zimmerer,Qi Gong,Wei Kang

【4】 Learning and Decision-Making with Data: Optimal Formulations and Phase Transitions
数据学习与决策:最优公式与相变
链接arxiv.org/abs/2109.0691
作者:M. Amine Bennouna,Bart P. G. Van Parys
机构:Bart P.G. Van Parys

预测|估计(2篇)

【1】 CAMul: Calibrated and Accurate Multi-view Time-Series Forecasting
CAMul:校准准确的多视图时间序列预测
链接arxiv.org/abs/2109.0743
作者:Harshavardhan Kamarthi,Lingkai Kong,Alexander Rodríguez,Chao Zhang,B. Aditya Prakash
机构:College of Computing, Georgia Institute of Technology
备注:16 pages, 4 figures

【2】 Multi View Spatial-Temporal Model for Travel Time Estimation
行程时间估计的多视点时空模型
链接arxiv.org/abs/2109.0740
作者:ZiChuan Liu,Zhaoyang Wu,Meng Wang
机构:Wuhan University of Technology, Wuhan, China, East China Normal University, Shanghai, China, Sun Yat-sen University, Guangzhou, China

其他神经网络|深度学习|模型|建模(12篇)

【1】 Challenges in Detoxifying Language Models
语言模型解毒面临的挑战
链接arxiv.org/abs/2109.0744
作者:Johannes Welbl,Amelia Glaese,Jonathan Uesato,Sumanth Dathathri,John Mellor,Lisa Anne Hendricks,Kirsty Anderson,Pushmeet Kohli,Ben Coppin,Po-Sen Huang
机构:DeepMind
备注:23 pages, 6 figures, published in Findings of EMNLP 2021

【2】 Can one hear the shape of a neural network?: Snooping the GPU via Magnetic Side Channel
人们能听到神经网络的形状吗?:通过磁侧通道窥探GPU
链接arxiv.org/abs/2109.0739
作者:Henrique Teles Maia,Chang Xiao,Dingzeyu Li,Eitan Grinspun,Changxi Zheng
机构:Adobe Research, Columbia University &, University of Toronto
备注:14 pages, accepted to USENIX Security 2022

【3】 Self-learn to Explain Siamese Networks Robustly
自学强势解读暹罗网络
链接arxiv.org/abs/2109.0737
作者:Chao Chen,Yifan Shen,Guixiang Ma,Xiangnan Kong,Srinivas Rangarajan,Xi Zhang,Sihong Xie
机构:∗Computer Science and Engineering Dept, Lehigh University †Laboratory of Trustworthy Distributed Computing and Service (MoE), BUPT, ‡University of
备注:Accepted to ICDM 2021

【4】 Cross-lingual Transfer of Monolingual Models
单语模型的跨语言迁移
链接arxiv.org/abs/2109.0734
作者:Evangelia Gogoulou,Ariel Ekgren,Tim Isbister,Magnus Sahlgren
机构:RISE, KTH, Peltarion

【5】 PoWareMatch: a Quality-aware Deep Learning Approach to Improve Human Schema Matching
PoWareMatch:一种改进人类模式匹配的质量感知深度学习方法
链接arxiv.org/abs/2109.0732
作者:Roee Shraga,Avigdor Gal
备注:Technical report of the paper {\sf PoWareMatch}: a Quality-aware Deep Learning Approach to Improve Human Schema Matching, accepted to ACM Journal of Data and Information Quality (JDIQ), Special Issue on Deep Learning for Data Quality

【6】 End-to-End Learning of Flowchart Grounded Task-Oriented Dialogs
基于流程图的任务型对话的端到端学习
链接arxiv.org/abs/2109.0726
作者:Dinesh Raghu,Shantanu Agarwal,Sachindra Joshi,Mausam
机构: Indian Institute of Technology, New Delhi, India, IBM Research, New Delhi, India
备注:D. Raghu and S.Agarwal contributed equally to this work

【7】 Integrating Sensing and Communication in Cellular Networks via NR Sidelink
通过NR侧链路集成蜂窝网络中的传感和通信
链接arxiv.org/abs/2109.0725
作者:Dariush Salami,Ramin Hasibi,Stefano Savazzi,Tom Michoel,Stephan Sigg
机构: Michoel are with the Department of Informatics at theUniversity of Bergen
备注:The paper is submitted to JSAC and it is still under review

【8】 Learning Mathematical Properties of Integers
学习整数的数学性质
链接arxiv.org/abs/2109.0723
作者:Maria Ryskina,Kevin Knight
机构:Language Technologies Institute, Carnegie Mellon University, DiDi Labs
备注:BlackboxNLP 2021

【9】 Automatic Symmetry Discovery with Lie Algebra Convolutional Network
基于李代数卷积网络的对称性自动发现
链接arxiv.org/abs/2109.0710
作者:Nima Dehmamy,Robin Walters,Yanchen Liu,Dashun Wang,Rose Yu
机构:Northwestern University, Northeastern University, University of California San Diego

【10】 Building Accurate Simple Models with Multihop
用多跳技术构建精确的简单模型
链接arxiv.org/abs/2109.0696
作者:Amit Dhurandhar,Tejaswini Pedapati

【11】 Quantitative reconstruction of defects in multi-layered bonded composites using fully convolutional network-based ultrasonic inversion
基于全卷积网络超声反演的多层粘结复合材料缺陷定量重建
链接arxiv.org/abs/2109.0728
作者:Jing Rao,Fangshu Yang,Huadong Mo,Stefan Kollmannsberger,Ernst Rank

【12】 Non-linear Independent Dual System (NIDS) for Discretization-independent Surrogate Modeling over Complex Geometries
复杂几何上离散化独立代理建模的非线性独立对偶系统(NIDS)
链接arxiv.org/abs/2109.0701
作者:James Duvall,Karthik Duraisamy,Shaowu Pan
机构:University of Michigan, Ann Arbor, MI

其他(24篇)

【1】 When Does Translation Require Context? A Data-driven, Multilingual Exploration
翻译什么时候需要语境?数据驱动的多语言探索
链接arxiv.org/abs/2109.0744
作者:Kayo Yin,Patrick Fernandes,André F. T. Martins,Graham Neubig
机构:Language Technologies Institute, Carnegie Mellon University, Pittsburgh, PA, Instituto Superior Técnico & LUMLIS (Lisbon ELLIS Unit), Lisbon, Portugal, Instituto de Telecomunicações, Lisbon, Portugal, Unbabel, Lisbon, Portugal

【2】 Should We Be Pre-training? An Argument for End-task Aware Training as an Alternative
我们是不是应该做好训练前的准备工作?将结束任务意识训练作为替代方案的论点
链接arxiv.org/abs/2109.0743
作者:Lucio M. Dery,Paul Michel,Ameet Talwalkar,Graham Neubig
机构: Carnegie Mellon University, ENS PSL University, Determined AI
备注:16 pages, 4 figures

【3】 Union: A Unified HW-SW Co-Design Ecosystem in MLIR for Evaluating Tensor Operations on Spatial Accelerators
UNION:用于评估空间加速器张量操作的MLIR中统一的软硬件协同设计生态系统
链接arxiv.org/abs/2109.0741
作者:Geonhwa Jeong,Gokcen Kestor,Prasanth Chatarasi,Angshuman Parashar,Po-An Tsai,Sivasankaran Rajamanickam,Roberto Gioiosa,Tushar Krishna
机构:Georgia Institute of Technology
备注:This paper is accepted to PACT 2021

【4】 Modular Neural Ordinary Differential Equations
模神经常微分方程
链接arxiv.org/abs/2109.0735
作者:Max Zhu,Prof. P Lio,Jacob Moss
机构:University of Cambridge
备注:4 pages

【5】 Comparing decision mining approaches with regard to the meaningfulness of their results
比较决策挖掘方法结果的意义
链接arxiv.org/abs/2109.0733
作者:Beate Scheibel,Stefanie Rinderle-Ma
机构:Research Group Workflow Systems and Technology, Computer Science, University of Vienna, Chair of Information Systems and Business Process Management, Department of Informatics, Technical University of Munich, Germany

【6】 Spline-PINN: Approaching PDEs without Data using Fast, Physics-Informed Hermite-Spline CNNs
Spline-Pinn:用快速的物理信息Hermite-Spline CNN逼近无数据偏微分方程
链接arxiv.org/abs/2109.0714
作者:Nils Wandel,Michael Weinmann,Michael Neidlin,Reinhard Klein
机构: University of Bonn, Delft University of Technology, RWTH Aachen
备注:Submitted to AAAI 2022

【7】 Parallel Constraint-Driven Inductive Logic Programming
并行约束驱动的归纳逻辑程序设计
链接arxiv.org/abs/2109.0713
作者:Andrew Cropper,Oghenejokpeme Orhobor,Cristian Dinu,Rolf Morel
备注:Paper under review

【8】 What Does The User Want? Information Gain for Hierarchical Dialogue Policy Optimisation
用户想要什么?分层对话策略优化的信息增益
链接arxiv.org/abs/2109.0712
作者:Christian Geishauser,Songbo Hu,Hsien-chin Lin,Nurul Lubis,Michael Heck,Shutong Feng,Carel van Niekerk,Milica Gašić
机构: Gaˇsi´c 1 1Heinrich Heine University D¨usseldorf, Germany 2Department of Computer Science and Technology, University of Cambridge

【9】 F-CAM: Full Resolution CAM via Guided Parametric Upscaling
F-CAM:基于导引参数放大的全分辨率CAM
链接arxiv.org/abs/2109.0706
作者:Soufiane Belharbi,Aydin Sarraf,Marco Pedersoli,Ismail Ben Ayed,Luke McCaffrey,Eric Granger
机构: LIVIA, Dept. of Systems Engineering, École de technologie supérieure, Montreal, Canada, Goodman Cancer Research Centre, Dept. of Oncology, McGill University, Montreal, Canada, Ericsson, Montreal, Canada
备注:23pages, under review

【10】 Self-Training with Differentiable Teacher
差异化教师的自我训练
链接arxiv.org/abs/2109.0704
作者:Simiao Zuo,Yue Yu,Chen Liang,Haoming Jiang,Siawpeng Er,Chao Zhang,Tuo Zhao,Hongyuan Zha
机构:⋄Georgia Institute of Technology, □Amazon, †The Chinese University of Hong Kong, Shenzhen

【11】 Attention Is Indeed All You Need: Semantically Attention-Guided Decoding for Data-to-Text NLG
注意力确实是你所需要的:数据到文本NLG的语义注意力引导解码
链接arxiv.org/abs/2109.0704
作者:Juraj Juraska,Marilyn Walker
机构:Natural Language and Dialogue Systems Lab, University of California, Santa Cruz
备注:Accepted to INLG 2021

【12】 Avengers Ensemble! Improving Transferability of Authorship Obfuscation
复仇者联盟!提高署名模糊的可转移性
链接arxiv.org/abs/2109.0702
作者:Muhammad Haroon,Muhammad Fareed Zaffar,Padmini Srinivasan,Zubair Shafiq
机构:Avengers Ensemble! Improving Transferability, of Authorship Obfuscation
备注:Submitted to PETS 2021

【13】 Embedding Node Structural Role Identity Using Stress Majorization
利用应力优化嵌入节点结构角色标识
链接arxiv.org/abs/2109.0702
作者:Lili Wang,Chenghan Huang,Weicheng Ma,Ying Lu,Soroush Vosoughi
机构:Dartmouth College, Hanover, New Hampshire, USA, Millennium Management, LLC, New York, New York, USA, Stony Brook Univeristy, Stony Brook, New York, USA
备注:In CIKM 2021

【14】 Behavior of k-NN as an Instance-Based Explanation Method
k-NN作为一种基于实例的解释方法的行为
链接arxiv.org/abs/2109.0699
作者:Chhavi Yadav,Kamalika Chaudhuri
机构:UC San Diego

【15】 A Crawler Architecture for Harvesting the Clear, Social, and Dark Web for IoT-Related Cyber-Threat Intelligence
用于获取物联网相关网络威胁情报的清晰、社交和黑暗网络的爬虫架构
链接arxiv.org/abs/2109.0693
作者:Paris Koloveas,Thanasis Chantzios,Christos Tryfonopoulos,Spiros Skiadopoulos
机构:University of the Peloponnese, GR, Tripolis, Greece
备注:None

【16】 Disentangling Generative Factors of Physical Fields Using Variational Autoencoders
用变分自动编码器解开物理场的生成因子
链接arxiv.org/abs/2109.0739
作者:Christian Jacobsen,Karthik Duraisamy
机构:University of Michigan, Ann Arbor, MI

【17】 How to use KL-divergence to construct conjugate priors, with well-defined non-informative limits, for the multivariate Gaussian
如何利用KL散度构造具有定义明确的非信息性极限的多元高斯共轭先验
链接arxiv.org/abs/2109.0738
作者:Niko Brümmer
备注:10 pages

【18】 Distribution-free Contextual Dynamic Pricing
免分销上下文动态定价
链接arxiv.org/abs/2109.0734
作者:Yiyun Luo,Will Wei Sun,and Yufeng Liu
机构: Krannert School of Management, Purdue University, Department of Statistics and Operations Research

【19】 DeFungi: Direct Mycological Examination of Microscopic Fungi Images
脱菌:真菌显微图像的直接真菌检查
链接arxiv.org/abs/2109.0732
作者:Camilo Javier Pineda Sopo,Farshid Hajati,Soheila Gheisari
机构:College of Engineering and Science, Victoria University Sydney, Sydney, NSW , Australia

【20】 Testing Self-Organized Criticality Across the Main Sequence using Stellar Flares from TESS
利用TESS的恒星耀斑测试跨越主序的自组织临界性
链接arxiv.org/abs/2109.0701
作者:Adina D. Feinstein,Darryl Z. Seligman,Maximilian N. Günther,Fred C. Adams
机构:Department of Astronomy and Astrophysics, University of Chicago, Chicago, IL , USA, Department of the Geophysical Sciences, University of Chicago, Chicago, IL , USA, Maximilian N. G¨unther
备注:6 pages, 3 figures, Submitted to journal

【21】 Scalable Average Consensus with Compressed Communications
可扩展的平均共识,支持压缩通信
链接arxiv.org/abs/2109.0699
作者:Mohammad Taha Toghani,César A. Uribe
机构: several averageThe authors are with the Department of Electrical and Computer En-gineering, Rice University

【22】 Combining GEDI and Sentinel-2 for wall-to-wall mapping of tall and short crops
GEDI和Sentinel-2结合用于高矮作物的壁间作图
链接arxiv.org/abs/2109.0697
作者:Stefania Di Tommaso,Sherrie Wang,David B. Lobell
机构: Department of Earth System Science and Center on Food Security and the, Environment, Stanford University, Institute for Computational and Mathematical Engineering, Stanford University, Goldman School of Public Policy, University of California, Berkeley

【23】 Targeted Cross-Validation
定向交叉验证
链接arxiv.org/abs/2109.0694
作者:Jiawei Zhang,Jie Ding,Yuhong Yang
机构:School of Statistics, University of Minnesota, and

【24】 Reconstruction on Trees and Low-Degree Polynomials
树与低次多项式的重构
链接arxiv.org/abs/2109.0691
作者:Frederic Koehler,Elchanan Mossel
机构:Simons Institute, Part of this work was completed while at the Department of Mathematics, †Department of Mathematics and IDSS, Massachusetts Institute of Technology
备注:20 pages, comments welcome

机器翻译,仅供参考


访问arxivdaily.com获取含摘要速递,更有收藏、搜索等功能,涵盖CS|物理|数学|经济|统计|金融|生物|电气领域
同步公众号:arXiv每日学术速递,欢迎关注
发布于 2021-09-16 09:38