Assistant Professor of Computer Science
Education
- 2023, Doctoral Degree, Information Systems, Carnegie-Mellon University
- 2016, Master's Degree, Computer Science, University of Toronto
- 2015, Bachelor's Degree, Computer Engineering, University of Cincinnati
Biography
Dr. Yue Zhao is an Assistant Professor of Computer Science at the University of Southern California and a faculty member of the USC Machine Learning Center. He leads the FORTIS Lab (Foundations Of Robust Trustworthy Intelligent Systems), where his research focuses on building robust, trustworthy, and scalable AI systems. His approach integrates four interconnected areas: developing robust foundations for anomaly and OOD detection; advancing graph-based learning methods; exploring transformative applications of large language and generative models; and democratizing AI through scalable, automated, and open-source ML systems. His work has widespread impacts in domains such as healthcare, finance, and political science. Dr. Zhao has authored over 60 papers in top-tier venues and is internationally recognized for his influential open-source projects—including PyOD, PyGOD, TDC, and TrustLLM—which collectively have earned over 20,000 GitHub stars (ranked top 0.00075% worldwide) and exceeded 30 million downloads. His projects have been widely adopted by leading organizations, including NASA, Morgan Stanley, Tesla, and the U.S. Senate Committee on Homeland Security & Governmental Affairs. Dr. Zhao's contributions have been honored by numerous awards, including the Capital One Research Award, Amazon Security AI Awards, AAAI New Faculty Highlights Award, Google Cloud Research Innovators Award, Norton Fellowship, Meta AI4AI Research Award, and the CMU Presidential Fellowship. He currently serves as an associate/action editor for ACM Transactions on AI for Science (TAIS), IEEE Transactions on Neural Networks and Learning Systems (TNNLS), and Journal of Data-centric Machine Learning Research (DMLR), and regularly serves as an area chair for leading ML conferences.
Research Summary
Robust AI Foundations: Anomaly and OOD Detection. The journey starts with developing foundational algorithms that reliably detect anomalies, outliers, and out-of-distribution (OOD) data, ensuring AI systems remain effective even under unexpected or adversarial scenarios.
Keywords: Anomaly Detection, Outlier Mining, OOD Detection, Robust AI
Graph-based Learning: Mining Complex Data Relationships. Building on robust detection foundations, I apply these principles to graph data, developing specialized algorithms for node-level outlier detection, open-set learning, and graph anomaly detection. This allows effective analysis of complex, relational information pervasive in social networks, finance, and security domains.
Keywords: Graph Learning, Node-level Outlier Detection, Graph Anomaly Detection, Open-set Learning
Transformative Applications: Large Language and Generative Models. Leveraging robust foundations and insights from complex relational data, I employ large language models (LLMs) and generative AI methods to tackle pressing societal challenges, including political forecasting and drug discovery.
Keywords: LLMs, Generative AI, Political Forecasting, Drug Discovery
Open-source and Scalable Systems: Democratizing AI Impact. Finally, to ensure these advanced methods are widely accessible, I build scalable, automated, and open-source frameworks such as PyOD (25M+ downloads, used by NASA, Tesla, and others). Collectively, my open-source projects have earned 20,000+ GitHub stars, ranking #750 worldwide, enabling reproducible and large-scale deployment of AI innovations.
Keywords: Automated ML, Open-source AI, Distributed Systems, Federated Learning, Scalability
Awards
- 2024 Amazon Amazon Research Awards
- 2024 Google Google Cloud Research Innovators
- 2024 Capital One Research Awards
