Cho-Jui Hsieh

Cho-Jui Hsieh is an Associate Professor of Computer Science at the University of California, Los Angeles (UCLA) in the Henry Samueli School of Engineering and Applied Science. He also serves as a research scientist at Google. Previously, he was an Assistant Professor at UC Davis Computer Science and Statistics for three years and a visiting scholar at Google since summer 2018.

Dr. Hsieh received his Ph.D. from UT Austin under the supervision of Prof. Inderjit Dhillon and his Master's degree from National Taiwan University under Prof. Chih-Jen Lin.

His research focuses on developing new algorithms and optimization techniques for large-scale machine learning problems. His current work emphasizes improving model size, training speed, prediction speed, and robustness of deep learning models. He is particularly known for his contributions to neural network verification, adversarial robustness, and optimization algorithms for machine learning. Dr. Hsieh has co-authored the book "Adversarial Robustness for Machine Learning" with Dr. Pin-Yu Chen.

His research trends show a strong emphasis on formal verification of deep neural networks, robustness certification, and optimization techniques for both traditional machine learning models and large language models. Recent work explores multimodal learning, neural network verification with branch-and-bound techniques, and certified training approaches.

• Frontiers of Science Award (2023)
• Okawa Research Award (2021)
• Outstanding Paper Award, ICLR (2021)
• Google Research Scholar Award (2021)
• NSF Career Award (2021)
• Samsung AI Researcher of the Year (2020)

Dr. Hsieh has supervised numerous PhD students, both current and former, many of whom have gone on to become assistant professors at prestigious institutions or work at leading AI companies like OpenAI, Meta, and Mistral AI. His research has been supported by various grants including the NSF Career Award. He leads a research group focused on machine learning algorithms and has developed several widely-used software tools including Distributed Kernel SVM and Asynchronous Kernel SVM. His current research directions include advancing neural network verification techniques, improving the robustness of large language models, and developing more efficient training algorithms for deep learning models.