Difference between revisions of "User:Snemoto"
From REU@MU
Line 29: | Line 29: | ||
**[https://dl.acm.org/doi/10.1145/3319535.3354209 Latent Backdoor Attacks on Deep Neural Networks] | **[https://dl.acm.org/doi/10.1145/3319535.3354209 Latent Backdoor Attacks on Deep Neural Networks] | ||
**[https://ieeexplore.ieee.org/abstract/document/8844607?casa_token=Vuy2khB6fo4AAAAA:5yHwrbSaHy_pTaLS0_poE87Ff8-htRLSqHOngkfFUVnE11AlBMFU-wCogOyMj3P3SJlGdn9R Membership Inference Attacks against Adversarially Robust Deep Learning Models] | **[https://ieeexplore.ieee.org/abstract/document/8844607?casa_token=Vuy2khB6fo4AAAAA:5yHwrbSaHy_pTaLS0_poE87Ff8-htRLSqHOngkfFUVnE11AlBMFU-wCogOyMj3P3SJlGdn9R Membership Inference Attacks against Adversarially Robust Deep Learning Models] | ||
− | **Seeing isn’t Believing: Towards More Robust Adversarial Attack Against Real World Object Detectors | + | **[https://dl.acm.org/doi/10.1145/3319535.3354259 Seeing isn’t Believing: Towards More Robust Adversarial Attack Against Real World Object Detectors] |
+ | **[https://www.computer.org/csdl/proceedings-article/sp/2019/666000a530/19skfH8dcqc Neural Cleanse: Identifying and Mitigating Backdoor Attacks in Neural Networks] |
Revision as of 20:00, 12 June 2020
Shota Nemoto
University: Case Western Reserve University
Currently pursuing a Bachelor's degree in Computer Science
Currently working with Dr. Perouli on the project: Identifying Appropriate Machine Learning Models for Multi Robot Secure Coordination in a Healthcare Facility.
Weekly Logs
Week 1
- Read papers on neural network inversion and HopSkipJump attacks.
- Read abstracts for HumptyDumpty and MemGuard papers
- Attended Orientation
- Filled out pre-REU survey
- Reviewed Python skills
- Learned basics of Pandas and DataFrame manipulation
- Learned basics of creating and evaluating machine learning models
- Learned basic security considerations for applications
Week 2
- Attended Responsible Conduct of Research Training
- Attended Technical Writing instruction session
- Completed CITI Modules
- Read papers:
- Read introductions and conclusions for papers:
- Latent Backdoor Attacks on Deep Neural Networks
- Membership Inference Attacks against Adversarially Robust Deep Learning Models
- Seeing isn’t Believing: Towards More Robust Adversarial Attack Against Real World Object Detectors
- Neural Cleanse: Identifying and Mitigating Backdoor Attacks in Neural Networks