PMLR

论文阅读“A probabilistic framework for multi-view feature learning with many-to-many associations via...

ShimodairaH.Aprobabilisticframeworkformulti-viewfeaturelearningwithmany-to-manyassociationsvianeuralnetworks[C]//InternationalConferenceonMachineLearning.PMLR

掉了西红柿皮_Kee·2024-02-01 11:54

Adversarial Attack on Graph Structured Data（2018 PMLR）

AdversarialAttackonGraphStructuredData----《图结构数据的对抗攻击》摘要基于图结构的深度学习已经在各种应用中显示出令人兴奋的结果。然而，与图像或文本对抗攻击和防御的大量研究工作相比，此类模型的鲁棒性却很少受到关注。在本文中，我们重点关注通过修改数据组合结构来欺骗深度学习模型的对抗性攻击。我们首先提出一种基于强化学习的攻击方法，该方法学习可泛化的攻击策略，

今我来思雨霏霏_JYF·2023-11-27 02:23

阅读记录【PMLR2023】The Aggregation–Heterogeneity Trade-off in Federated Learning

TheAggregation–HeterogeneityTrade-offinFederatedLearningLink：https://proceedings.mlr.press/v195/zhao23b.htmlAbstract机器学习的传统观点认为，训练模型的数据越多，模型的性能就越好。因此，人们开发了多种联邦学习方法来聚合尽可能多的本地样本。与这种观点相反，本文表明，在存在异构性的情况下，

furoto_·2023-11-26 08:57

理解训练深度前馈神经网络的难度【PMLR 2010】

论文地址：Excellent-Paper-For-Daily-Reading/summarizeatmain类别：综述时间：2023/11/03摘要这篇论文比较久了，但仍能从里面获得一些收获，论文主要是讨论并研究了不同的非线性激活函数的影响，sigmoid函数它的非零均值会在Hessian中诱发重要的奇异值，很容易导致在隐藏层中达到饱和区域，且也证实了sigmoid激活函数在随机初始化的深度网络并

夏天是冰红茶·2023-11-03 23:05

[论文阅读72]Parameter-Efficient Transfer Learning for NLP

1.基本信息题目论文作者与单位来源年份Parameter-EfficientTransferLearningforNLPNeilHoulsby等GoogleResearch，雅盖隆大学-波兰PMLR2019HoulsbyN

happyprince·2023-10-21 18:15

【论文阅读笔记】CRFL: Certifiably Robust Federated Learning against Backdoor Attacks

会议：PMLR2021[2106.08283]CRFL:CertifiablyRobustFederatedLearningagainstBackdoorAttacks(arxiv.org)问题：现有的防御算法缺乏健壮性创新

MiaZX·2023-06-17 19:50

BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Genera

Bootstrappinglanguage-imagepre-trainingforunifiedvision-languageunderstandingandgeneration[C]//InternationalConferenceonMachineLearning.PMLR

游不动的鱼-learning·2023-06-07 23:44

[TRPO] Trust Region Policy Optimization

SchulmanJ,LevineS,AbbeelP,etal.Trustregionpolicyoptimization[C]//Internationalconferenceonmachinelearning.PMLR

超级超级小天才·2023-03-30 23:55

ViLT：不用卷积/区域特征监督信号的视觉-语言Transformer(速读版)

InternationalConferenceonMachineLearning.PMLR,2021.源码：https://github.com/dandelin/vilt视觉

Civisky·2022-12-25 22:47

《论文阅读》ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision

Vision-and-LanguageTransformerWithoutConvolutionorRegionSupervision简介解决的问题模型构架数据集实验分析创新点出版：Proceedingsofthe38thInternationalConferenceonMachineLearning,PMLR

365JHWZGo·2022-12-25 22:15

分布式强化学习方法汇总

Scalabledistributeddeep-rlwithimportanceweightedactor-learnerarchitectures[C]//InternationalConferenceonMachineLearning.PMLR

Barry Yang·2022-12-23 13:04

ViLT解读记录

Vision-and-LanguageTransformerWithoutConvolutionorRegionSupervision视频讲解：ViLT论文精读论文下载：https://arxiv.org/pdf/2102.03334.pdf代码地址：https://github.com/dandelin/vilt论文投稿：PMLR2021ProceedingsofMachineLe

如果你也听说~·2022-12-20 17:04

clip-summary

LearningTransferableVisualModelsFromNaturalLanguageSupervision视频讲解：CLIP论文精读论文下载：https://openai.com/blog/clip/代码地址：https://github.com/openai/CLIP论文投稿：PMLR2021ProceedingsofMachineLearningResearch

如果你也听说~·2022-12-09 17:47

Meta-World：多任务、持续学习、终身学习、元学习、强化学习的基准和评估

1.背景论文链接，斯坦福大学、加州大学伯克利分校、哥伦比亚大学、南加州大学、谷歌联合完成，发表于PMLR2020paperwithcode网页链接项目主页链接源代码链接2.基准相关的持续学习论文LifelongPolicyGradientLearningofFactoredPoliciesforFasterTrainingWithoutForgettingLifelongMachineLearni

阿航626·2022-12-03 13:22

[PMLR 2017] On calibration of modern neural networks

ContentsIntroductionCalibrationMetricsObservingMiscalibrationCalibrationMethodsCalibratingBinaryModelsExtensiontoMulticlassModelsResultsReferencesIntroduction在实际的决策系统中，分类模型不仅需要尽量给出准确的预测，还需要能告诉我们它给出的预测

连理o·2022-12-02 17:29

论文(五)：Cluster-to-Conquer: A Framework for End-to-End Multi-Instance Learning for Whole Slide Image

Cluster-to-Conquer:AFrameworkforEnd-to-EndMulti-InstanceLearningforWholeSlideImageClassification--PMLR

小王不叫小王叭·2022-11-20 05:34

论文阅读“Contrastive multi-view representation learning on graphs”

KhasahmadiAH.Contrastivemulti-viewrepresentationlearningongraphs[C]//InternationalConferenceonMachineLearning.PMLR

掉了西红柿皮_Kee·2022-07-28 19:04

Transfer Learning via Learning to Transfer

后来一篇《TransferLearningviaLearningtoTransfer》被PMLR2018接受。

gdtop818·2020-06-28 20:08

ICML2019

去pmlr下载icml的bibliography.bib用如下代码parse.py预处理成latex段落#!

颹蕭蕭·2019-06-04 11:59

推荐频道