Dataset Security for Machine Learning: Data Poisoning, Backdoor Attacks, and Defenses
Author:
Affiliation:
1. University of Maryland, College Park, MD, USA
2. Massachusetts Institute of Technology, Cambridge, MA, USA
3. University of Illinois, Champaign, IL, USA
4. University of California, Berkeley, CA, USA
Funder
Defense Advanced Research Projects Agency
QED4RML
D3M programs
National Science Foundation
Amazon Research Award Program
Berkeley DeepDrive
Facebook PhD Fellowship
ONR MURI Program
Alfred P. Sloan Foundation
Publisher
Institute of Electrical and Electronics Engineers (IEEE)
Subject
Applied Mathematics,Artificial Intelligence,Computational Theory and Mathematics,Computer Vision and Pattern Recognition,Software
Link
https://ieeexplore.ieee.org/ielam/34/10008914/9743317-aam.pdf
Reference205 articles.
1. Influence functions in deep learning are fragile;basu,2020
2. Stronger data poisoning attacks break data sanitization defenses;koh,2018
3. Efficient Label Contamination Attacks Against Black-Box Learning Models
4. Auditing differentially private machine learning: How private is private SGD?;jagielski;in Proc Adv Neural Informat Process Syst,2020
5. Support vector machines under adversarial label noise;biggio;Proc Asian Conf Mach Learn,2011
Cited by 51 articles. 订阅此论文施引文献 订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献
1. Toward a Critical Evaluation of Robustness for Deep Learning Backdoor Countermeasures;IEEE Transactions on Information Forensics and Security;2024
2. Backdoor attacks and defenses in federated learning: Survey, challenges and future research directions;Engineering Applications of Artificial Intelligence;2024-01
3. Learning a robust foundation model against clean-label data poisoning attacks at downstream tasks;Neural Networks;2024-01
4. LearnChain: Transparent and cooperative reinforcement learning on Blockchain;Future Generation Computer Systems;2024-01
5. Security of NVMe Offloaded Data in Large-Scale Machine Learning;Computer Security – ESORICS 2023;2024
1.学者识别学者识别
2.学术分析学术分析
3.人才评估人才评估
"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370
www.globalauthorid.com
TOP
Copyright © 2019-2024 北京同舟云网络信息技术有限公司 京公网安备11010802033243号 京ICP备18003416号-3