Amazon Fellows and faculty-led projects advance innovations in machine learning and artificial intelligence
The Amazon–Virginia Tech Initiative for Efficient and Robust Machine Learning has announced support for two Amazon Fellows and four innovative research projects led by Virginia Tech faculty that further the initiative’s mission of advancing new directions in machine learning.
Funded by Amazon, housed in the College of Engineering, and directed by researchers at the Sanghani Center for Artificial Intelligence and Data Analytics on Virginia Tech’s campus in Blacksburg and at the Innovation Campus in Alexandria, the initiative was launched in March to support student and faculty-led development and implementation of innovative approaches to robust machine learning — such as ensuring that algorithms and models are resistant to errors and adversaries — that could address worldwide industry-focused problems.
An open call for student fellowship nominations and research projects went out concurrently across the Virginia Tech campuses. The initiative’s advisory committee, comprised of Virginia Tech faculty and Amazon researchers, selected two Amazon Fellows from among 11 nominations and four faculty award recipients from 14 submissions.
“Our inaugural cohort of fellows and faculty-led projects showcases the breadth of machine learning research happening at Virginia Tech,” said Naren Ramakrishnan, the Thomas L. Phillips Professor of Engineering and director of the Amazon-Virginia Tech Initiative. “The areas represented include federated learning, meta-learning, leakage from machine learning models, and conversational interfaces.
“This research will not only contribute to new algorithmic advances, but also study issues pertaining to practical and safe deployment of machine learning,” said Ramakrishnan who also directs the Sanghani Center for Artificial Intelligence and Data Analytics. “We are very excited that the partnership between Amazon and Virginia Tech has enabled these projects.”
"The talent and depth of scientific knowledge at Virginia Tech is reflected in the high-quality research proposals and Ph.D. student fellowship applications we have received,” said Prem Natarajan, vice president of Alexa AI. “I am excited about the new insights and advances in robust machine learning that will result from the work of the faculty and students who are contributing to this initiative."
The Amazon Fellows are:
- Qing Guo, a Ph.D. student in the Department of Statistics, whose research in machine learning covers nonparametric mutual information estimation with contrastive learning techniques; optimal Bayesian experimental design for both static and sequential models; meta-learning based on information-theoretic generalization theory; and reasoning for conversational search and recommendation. Her advisor is Xinwei Deng.
- Yi Zeng, a Ph.D. student in the Bradley Department of Electrical and Computer Engineering, whose research interests focus on assessing potential risks as artificial intellingence (AI) is increasingly used to support essential societal tasks, such as health care, business activities, financial services, and scientific research, and developing practical and effective countermeasures for the safe deployment of AI. His advisor is Ruoxi Jia.
Principal investigators and co-principal investigators, respectively, for funded projects are:
- Peng Gao, assistant professor, Department of Computer Science, and Ruoxi Jia, assistant professor, Bradley Department of Electrical and Computer Engineering, and faculty at the Sanghani Center — “Platform-Agnostic Privacy Leakage Monitoring for Machine Learning Models.” The project proposes new platform-agnostic privacy leakage detection methods by identifying self-similar, low-utility model queries, and a stream-based system architecture that enables real-time privacy leakage monitoring and detection.
- Jia and Yalin Sagduyu, a research professor in the Intelligent Systems Division of the Virginia Tech National Security Institute — “FEDGUARD: Safeguard Federated Learning Systems against Backdoor Attacks.” Successful completion of this project will provide key enabling technologies for secure federated learning and accelerate its adoption in security-sensitive applications such as digital assistant systems.
- Ismini Lourentzou, assistant professor, Department of Computer Science, and faculty at the Sanghani Center — “Toward Unified Multimodal Conversational Embodied Agents.” The project will develop a general-purpose embodied agent that can understand instructions, interact with humans, predict human beliefs, and reason to complete a broad range of tasks.
- Walid Saad, professor, Bradley Department of Electrical and Computer Engineering — “Green, Efficient, and Scalable Federated Learning over Resource-Constrained Devices and Systems.” The research advances techniques from machine learning, wireless communications, game theory, and mean-field theory to support federated learning over real-world wireless systems and resource-constrained devices.