Improved Protocols for Privacy-Preserving Machine Learning on the GPU; Industry Adaptation of Privacy-Preserving Computation

Blindenbach, Jacob, School of Engineering and Applied Science, University of Virginia
Seabrook, Bryn, EN-Engineering and Society, University of Virginia
Tian, Yuan, EN-Comp Science Dept, University of Virginia

Machine learning and artificial intelligence power sophisticated technologies like personal assistance, Google search, self-driving cars, etc. These technologies require millions of user data points to train and finetune the algorithms. Ensuring the security and privacy of this user data is critical to maintain the protection of user data rights. Furthermore, as user data becomes more irreplaceable, for example, personal genomic data or finger print data, the consequences of data violations and data breaches grow. This portfolio examines privacy-preserving computation which is utilized in privacy-preserving machine learning to provably guarantee privacy and security of user data used in machine learning. The technical research for this portfolio describes efficient cryptographic protocols for privacy-preserving machine learning while the STS portion of this portfolio focuses more broadly on privacy-preserving computation. The STS research analyzes the question: what are the primary causes for the lack of industry adaptation of privacy-preserving computation?

BS (Bachelor of Science)
Privacy-Preserving Machine Learning, Hardware Acceleration, Industry Adaptation

School of Engineering and Applied Science
Bachelor of Science in Computer Science
Technical Advisor: Yuan Tian
STS Advisor: Bryn Seabrook

Issued Date: