同步wx订阅号(arXiv每日论文速递),支持后台回复'search 关键词'查询相关的最新论文。有些许帮助的话,麻烦关注一下哦(* ̄rǒ ̄)
cs.LG 方向,今日共计72篇
[cs.LG]:
【1】 When to Trust Your Model: Model-Based Policy Optimization
标题:何时信任模型:基于模型的策略优化
作者: Michael Janner, Sergey Levine
链接:https://arxiv.org/abs/1906.08253
【2】 Provable Gradient Variance Guarantees for Black-Box Variational Inference
标题:黑箱变分推理的可证明梯度方差保证
作者: Justin Domke
链接:https://arxiv.org/abs/1906.08241
【3】 Evaluating Protein Transfer Learning with TAPE
标题:用胶带评价蛋白质转移学习
作者: Roshan Rao, Yun S. Song
备注:20 pages, 4 figures
链接:https://arxiv.org/abs/1906.08230
【4】 Unsupervised State Representation Learning in Atari
标题:无监督状态表示学习在Atari中的应用
作者: Ankesh Anand, R Devon Hjelm
链接:https://arxiv.org/abs/1906.08226
【5】 Clustering with Fairness Constraints: A Flexible and Scalable Approach
标题:具有公平性约束的集群:一种灵活且可扩展的方法
作者: Imtiaz Masud Ziko, Ismail Ben Ayed
链接:https://arxiv.org/abs/1906.08207
【6】 Control What You Can: Intrinsically Motivated Task-Planning Agent
标题:控制所能做的:本质上有动机的任务-计划代理
作者: Sebastian Blaes, Georg Martius
链接:https://arxiv.org/abs/1906.08190
【7】 QXplore: Q-learning Exploration by Maximizing Temporal Difference Error
标题:QXplore:通过最大化时差误差进行Q学习探索
作者: Riley Simmons-Edler, Daniel Lee
链接:https://arxiv.org/abs/1906.08189
【8】 PABO: Pseudo Agent-Based Multi-Objective Bayesian Hyperparameter Optimization for Efficient Neural Accelerator Design
标题:基于伪Agent的高效神经加速器多目标贝叶斯超参数优化设计
作者: Maryam Parsa, Kaushik Roy
链接:https://arxiv.org/abs/1906.08167
【9】 BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learning
标题:BatchBALD:用于深度贝叶斯主动学习的高效且多样化的批量获取
作者: Andreas Kirsch, Yarin Gal
链接:https://arxiv.org/abs/1906.08158
【10】 Learning Disentangled Representations of Timbre and Pitch for Musical Instrument Sounds Using Gaussian Mixture Variational Autoencoders
标题:使用高斯混合变分自动编码器学习乐器声音的音色和音高的解缠表示
作者: Yin-Jyun Luo, Dorien Herremans
链接:https://arxiv.org/abs/1906.08152
【11】 Generative Restricted Kernel Machines
标题:生成受限核机器
作者: Arun Pandey, Johan A. K. Suykens
链接:https://arxiv.org/abs/1906.08144
【12】 Efficient Algorithms for Set-Valued Prediction in Multi-Class Classification
标题:多类分类中集值预测的有效算法
作者: Thomas Mortier, Willem Waegeman
链接:https://arxiv.org/abs/1906.08129
【13】 Learning in Restless Multi-Armed Bandits via Adaptive Arm Sequencing Rules
标题:基于自适应ARM排序规则的无休止多臂带学习
作者: Tomer Gafni, Kobi Cohen
备注:A short version of this paper was presented at IEEE International Symposium on Information Theory (ISIT) 2018
链接:https://arxiv.org/abs/1906.08120
【14】 Wasserstein Adversarial Imitation Learning
标题:瓦瑟斯坦对抗性模仿学习
作者: Huang Xiao, Thai Hong Linh
链接:https://arxiv.org/abs/1906.08113
【15】 Constrained Bilinear Factorization Multi-view Subspace Clustering
标题:约束双线性分解多视图子空间聚类
作者: Qinghai Zheng, Xiuyi Jia
链接:https://arxiv.org/abs/1906.08107
【16】 Transfer NAS: Knowledge Transfer between Search Spaces with Transformer Agents
标题:Transfer NAS:使用Transform Agent在搜索空间之间进行知识转移
作者: Zalán Borsos, Andrea Gesmundo
链接:https://arxiv.org/abs/1906.08102
【17】 Automatic Source Code Summarization with Extended Tree-LSTM
标题:基于扩展树的源代码自动总结-LSTM
作者: Yusuke Shido, Tadayuki Matsumura
备注:IJCNN 2019, to appear
链接:https://arxiv.org/abs/1906.08094
【18】 LIA: Latently Invertible Autoencoder with Adversarial Learning
标题:LIA:具有对抗性学习的潜在可逆自动编码器
作者: Jiapeng Zhu, Bo Zhang
链接:https://arxiv.org/abs/1906.08090
【19】 Online Heterogeneous Mixture Learning for Big Data
标题:面向大数据的在线异构混合学习
作者: Kazuki Seshimo, Yamane Satoshi
链接:https://arxiv.org/abs/1906.08068
【20】 Learning Directed Graphical Models from Gaussian Data
标题:从高斯数据中学习有向图形模型
作者: Katherine Fitch
链接:https://arxiv.org/abs/1906.08050
【21】 Barron Spaces and the Compositional Function Spaces for Neural Network Models
标题:Barron空间与神经网络模型的复合函数空间
作者: Weinan E, Lei Wu
链接:https://arxiv.org/abs/1906.08039
【22】 Disentangling feature and lazy learning in deep neural networks: an empirical study
标题:深神经网络中的解缠特征与懒惰学习:一项实证研究
作者: Mario Geiger, Matthieu Wyart
链接:https://arxiv.org/abs/1906.08034
【23】 XNAS: Neural Architecture Search with Expert Advice
标题:XNAs:带专家建议的神经结构搜索
作者: Niv Nayman, Lihi Zelnik-Manor
链接:https://arxiv.org/abs/1906.08031
【24】 Adaptive Temporal-Difference Learning for Policy Evaluation with Per-State Uncertainty Estimates
标题:基于每状态不确定性估计的自适应时差学习策略评估
作者: Hugo Penedones, Gergely Neu
链接:https://arxiv.org/abs/1906.07987
【25】 A unified view on differential privacy and robustness to adversarial examples
标题:对不同隐私和对抗性例子的鲁棒性的统一看法
作者: Rafael Pinot, Jamal Atif
链接:https://arxiv.org/abs/1906.07982
【26】 Batch Active Learning Using Determinantal Point Processes
标题:基于行列点过程的批量主动学习
作者: Erdem Bıyık, Dorsa Sadigh
备注:Submitted to NeurIPS 2019
链接:https://arxiv.org/abs/1906.07975
【27】 SemanticAdv: Generating Adversarial Examples via Attribute-conditional Image Editing
标题:语义添加:通过属性条件图像编辑生成对抗性示例
作者: Haonan Qiu, Bo Li
链接:https://arxiv.org/abs/1906.07927
【28】 Global Adversarial Attacks for Assessing Deep Learning Robustness
标题:评估深度学习健壮性的全局对抗性攻击
作者: Hanbin Hu, Peng Li
备注:Submitted to NeurIPS 2019
链接:https://arxiv.org/abs/1906.07920
【29】 Convergence of Adversarial Training in Overparametrized Networks
标题:过参数化网络中对抗性训练的收敛性
作者: Ruiqi Gao, Jason D. Lee
链接:https://arxiv.org/abs/1906.07916
【30】 Discovery of Physics from Data: Universal Laws and Discrepancy Models
标题:从数据中发现物理:普遍规律和差异模型
作者: Brian de Silva (1), (3) University of Washington Mechanical Engineering)
链接:https://arxiv.org/abs/1906.07906
【31】 Adversarial Task-Specific Privacy Preservation under Attribute Attack
标题:属性攻击下的对抗性任务专用隐私保护
作者: Han Zhao, Geoffrey J. Gordon
链接:https://arxiv.org/abs/1906.07902
【32】 Joint Pruning on Activations and Weights for Efficient Neural Networks
标题:有效神经网络激活与权值的联合修剪
作者: Qing Yang, Hai Li
链接:https://arxiv.org/abs/1906.07875
【33】 Adapting Behaviour via Intrinsic Reward: A Survey and Empirical Study
标题:通过内在奖励适应行为:一项调查与实证研究
作者: Cam Linke, Adam White
链接:https://arxiv.org/abs/1906.07865
【34】 Supervised Hierarchical Clustering with Exponential Linkage
标题:基于指数链的有监督层次聚类
作者: Nishant Yadav, Andrew McCallum
备注:Appears in ICML 2019
链接:https://arxiv.org/abs/1906.07859
【35】 Agnostic data debiasing through a local sanitizer learnt from an adversarial network approach
标题:从对抗性网络方法中学到的通过本地消毒剂消除不可知性数据
作者: Ulrich Aïvodji, Alain Tapp
链接:https://arxiv.org/abs/1906.07858
【36】 Deep Learning-Based Quantization of L-Values for Gray-Coded Modulation
标题:基于深度学习的灰色编码调制L-值量化
作者: Marius Arvinte, Ahmed H. Tewfik
备注:Submitted to IEEE Globecom 2019
链接:https://arxiv.org/abs/1906.07849
【37】 Gradient Dynamics of Shallow Univariate ReLU Networks
标题:浅一元Relu网络的梯度动力学
作者: Francis Williams, Joan Bruna
链接:https://arxiv.org/abs/1906.07842
【38】 RadGrad: Active learning with loss gradients
标题:RADGrad:丢失梯度的主动学习
作者: Paul Budnarain, Ilan Kogan
链接:https://arxiv.org/abs/1906.07838
【39】 Information matrices and generalization
标题:信息矩阵与推广
作者: Valentin Thomas, Nicolas Le Roux
链接:https://arxiv.org/abs/1906.07774
【40】 Poisoning Attacks with Generative Adversarial Nets
标题:用生成性对抗性网络进行中毒攻击
作者: Luis Muñoz-González, Emil C. Lupu
链接:https://arxiv.org/abs/1906.07773
【41】 On the Robustness of the Backdoor-based Watermarking in Deep Neural Networks
标题:基于后门的深神经网络水印的鲁棒性研究
作者: Masoumeh Shafieinejad, Florian Kerschbaum
链接:https://arxiv.org/abs/1906.07745
【42】 Predicting Patent Citations to measure Economic Impact of Scholarly Research
标题:预测专利引文以衡量学术研究的经济影响
作者: Abdul Rahman Shaikh, Hamed Alhoori
链接:https://arxiv.org/abs/1906.08244
【43】 XLNet: Generalized Autoregressive Pretraining for Language Understanding
标题:XLNet:语言理解的广义自回归预训练
作者: Zhilin Yang, Quoc V. Le
链接:https://arxiv.org/abs/1906.08237
【44】 PyRobot: An Open-source Robotics Framework for Research and Benchmarking
标题:PyRobot:一种用于研究和标杆的开源机器人框架
作者: Adithyavairavan Murali, Abhinav Gupta
链接:https://arxiv.org/abs/1906.08236
【45】 Local Bures-Wasserstein Transport: A Practical and Fast Mapping Approximation
标题:局部Bures-Wasserstein传输:一种实用的快速映射近似
作者: Andrés Hoyos-Idrobo
链接:https://arxiv.org/abs/1906.08227
【46】 Variational Gaussian Processes with Signature Covariances
标题:具有特征协方差的变分高斯过程
作者: Csaba Toth, Harald Oberhauser
链接:https://arxiv.org/abs/1906.08215
【47】 An Outlier-aware Consensus Protocol for Blockchain-based IoT Networks Using Hyperledger Fabric
标题:基于超分类结构的块链IoT网络异常点感知一致性协议
作者: Mehrdad Salimitari, Mainak Chatterjee
备注:This paper is submitted to IEEE GLOBECOM 2019. It is under review
链接:https://arxiv.org/abs/1906.08177
【48】 Automatic Model Parallelism for Deep Neural Networks with Compiler and Hardware Support
标题:具有编译器和硬件支持的深神经网络自动模型并行化
作者: Sanket Tavarageri, Bharat Kaul
链接:https://arxiv.org/abs/1906.08168
【49】 Pre-Training with Whole Word Masking for Chinese BERT
标题:汉语BERT的全词掩蔽预训练
作者: Yiming Cui, Guoping Hu
链接:https://arxiv.org/abs/1906.08101
【50】 Monocular 3D Object Detection and Box Fitting Trained End-to-End Using Intersection-over-Union Loss
标题:基于交叉点超过联合损失的单目三维目标检测和盒拟合训练
作者: Eskil Jörgensen, Fredrik Kahl
链接:https://arxiv.org/abs/1906.08070
【51】 Automated Computer Evaluation of Acute Ischemic Stroke and Large Vessel Occlusion
标题:急性缺血性卒中和大血管闭塞的计算机自动评估
作者: Jia You, Gilberto K.K. Leung
链接:https://arxiv.org/abs/1906.08059
【52】 Speech Recognition With No Speech Or With Noisy Speech Beyond English
标题:无语音或英语以外有噪声语音的语音识别
作者: Gautam Krishna, Ahmed H Tewfik
链接:https://arxiv.org/abs/1906.08045
【53】 Robust End to End Speaker Verification Using EEG
标题:基于EEG的健壮端到端说话人验证
作者: Yan Han, Ahmed H Tewfik
链接:https://arxiv.org/abs/1906.08044
【54】 Low-resource Deep Entity Resolution with Transfer and Active Learning
标题:具有转移和主动学习的低资源深度实体解析
作者: Jungo Kasai, Lucian Popa
备注:This paper is accepted by ACL 2019
链接:https://arxiv.org/abs/1906.08042
【55】 Cloud-based Image Classification Service Is Not Robust To Simple Transformations: A Forgotten Battlefield
标题:基于云的图像分类服务对简单的转换不是很健壮:一个被遗忘的战场
作者: Dou Goodman, Tao Wei
链接:https://arxiv.org/abs/1906.07997
【56】 Explanations can be manipulated and geometry is to blame
标题:解释可以被操纵,几何学是罪魁祸首。
作者: Ann-Kathrin Dombrowski, Pan Kessel
链接:https://arxiv.org/abs/1906.07983
【57】 Imbalanced Learning-based Automatic SAR Images Change Detection by Morphologically Supervised PCA-Net
标题:基于不平衡学习的基于形态监督PCA网的SAR图像变化自动检测
作者: Rongfang Wang, Mi Wang
链接:https://arxiv.org/abs/1906.07923
【58】 Multimodal Abstractive Summarization for How2 Videos
标题:HOW2视频的多模态抽象总结
作者: Shruti Palaskar, Florian Metze
备注:To appear in ACL 2019
链接:https://arxiv.org/abs/1906.07901
【59】 Brain correlates of task-load and dementia elucidation with tensor machine learning using oddball BCI paradigm
标题:基于奇球BCI范式的张量机学习与任务负荷和痴呆解释的脑相关性
作者: Tomasz M. Rutkowski, Mihoko Otake-Matsuura
备注:In ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 8578-8582, May 2019
链接:https://arxiv.org/abs/1906.07899
【60】 Learning with Partially Ordered Representations
标题:用偏序表示学习
作者: Jane Chandlee, Jonathan Rawski
备注:to appear in Proceedings of Mathematics of Language (ACL SIGMOL 2019)
链接:https://arxiv.org/abs/1906.07886
【61】 Semi-supervised Logistic Learning Based on Exponential Tilt Mixture Models
标题:基于指数倾斜混合模型的半监督Logistic学习
作者: Xinwei Zhang, Zhiqiang Tan
链接:https://arxiv.org/abs/1906.07882
【62】 Identification and Estimation of Hierarchical Latent Attribute Models
标题:层次潜在属性模型的辨识与估计
作者: Yuqi Gu, Gongjun Xu
链接:https://arxiv.org/abs/1906.07869
【63】 Stochastic Runge-Kutta Accelerates Langevin Monte Carlo and Beyond
标题:随机Runge-Kutta加速Langevin Monte Carlo及更多
作者: Xuechen Li, Murat A. Erdogdu
链接:https://arxiv.org/abs/1906.07868
【64】 Locally Accelerated Conditional Gradients
标题:局部加速条件梯度
作者: Alejandro Carderera, Sebastian Pokutta
链接:https://arxiv.org/abs/1906.07867
【65】 Multi-user Resource Control with Deep Reinforcement Learning in IoT Edge Computing
标题:物联网边缘计算中基于深度强化学习的多用户资源控制
作者: Lei Lei, Xianbin Wang
链接:https://arxiv.org/abs/1906.07860
【66】 Key Instance Selection for Unsupervised Video Object Segmentation
标题:无监督视频对象分割的关键实例选择
作者: Donghyeon Cho, Jiwon Kim
备注:Ranked 3rd on the leaderboard of 'Unsupervised DAVIS Challenge' (CVPR 2019)
链接:https://arxiv.org/abs/1906.07851
【67】 Dataless training of generative models for the inverse design of metasurfaces
标题:元表面逆向设计生成模型的无数据训练
作者: Jiaqi Jiang, Jonathan A. Fan
链接:https://arxiv.org/abs/1906.07843
【68】 A Static Analysis-based Cross-Architecture PerformancePrediction Using Machine Learning
标题:基于静态分析的机器学习跨结构性能预测
作者: Newsha Ardalani, Karu Sankaralingam
备注:Published at 2nd International Workshop on AI-assisted Design for Architecture Phoenix, AZ, June 22, 2019, colocated with ISCA
链接:https://arxiv.org/abs/1906.07840
【69】 Kernel quadrature with DPPs
标题:带DPP的核正交
作者: Ayoub Belhadji, Pierre Chainais
链接:https://arxiv.org/abs/1906.07832
【70】 Safe Testing
标题:安全测试
作者: Peter Grünwald, Wouter Koolen
链接:https://arxiv.org/abs/1906.07801
【71】 Cascaded Cross-Module Residual Learning towards Lightweight End-to-End Speech Coding
标题:面向轻量级端到端语音编码的级联跨模块剩余学习
作者: Kai Zhen, Minje Kim
备注:Accepted for publication in INTERSPEECH 2019
链接:https://arxiv.org/abs/1906.07769
【72】 The Limit Points of (Optimistic) Gradient Descent in Min-Max Optimization
标题:Min-Max优化中(乐观)梯度下降的极限点
作者: Constantinos Daskalakis, Ioannis Panageas
链接:https://arxiv.org/abs/1807.03907
翻译:腾讯翻译