27 lines
1.0 KiB
Markdown
27 lines
1.0 KiB
Markdown
*Security and privacy* are rapidly emerging as critical research areas in
|
|
computer science and beyond. Vulnerabilities in software are found and exploited
|
|
almost everyday, with grave consequences. Personal data today is aggregated at
|
|
large scales, increasing the risk of privacy violations or breaches. Finally,
|
|
*machine-learning* (ML) algorithms are seeing real-world applications in
|
|
critical sectors (e.g., health care, automation, and finance), but their
|
|
behavior in the presence of malicious adversaries is poorly understood.
|
|
|
|
This advanced topics class will cover recent techniques from the frontiers of
|
|
security and privacy research. Topics will be drawn from the following broad
|
|
areas, depending on student interest:
|
|
|
|
### Differential Privacy
|
|
- Basic properties and examples
|
|
- Advanced mechanisms
|
|
- Local differential privacy
|
|
|
|
### Adversarial Machine Learning
|
|
- Training-time attacks
|
|
- Test-time attacks
|
|
- Model-theft attacks
|
|
|
|
### Cryptographic Techniques
|
|
- Zero-knowledge proofs
|
|
- Secure multi-party computation
|
|
- Verifiable computation
|