Shahriar Noroozizadeh

Shahriar Noroozizadeh is a Ph.D. student at Carnegie Mellon University, jointly affiliated with the Machine Learning Department (School of Computer Science) and the Heinz College of Information Systems and Public Policy. Advised by Prof. George Chen (Heinz/MLD) and Prof. Jeremy Weiss (National Library of Medicine at NIH), he focuses on interpretable representation learning for temporal data in healthcare, particularly for patient risk progression modeling and hospital discharge delay prediction.

• Education: Ph.D. in progress (ML & Public Policy, CMU); M.S. in Machine Learning (2022) and Biomedical Engineering (2020) from CMU; B.A.Sc. in Engineering Physics (EECS specialization) from UBC.

His research bridges machine learning methodology with high-stakes decision-making systems in healthcare, emphasizing interpretability, contrastive learning, and multimodal representation. He has pioneered supervised contrastive frameworks that outperform traditional models in clinical outcome prediction and synthetic dataset recovery, while also exploring embodied AI and large language models for molecular design and robotics.

Recent work includes a patent-pending mRNA language model (mRNA-LM) for structure prediction and translation efficiency analysis, developed during his A.I. Research Scientist Internship at Sanofi. His publications span venues like NeurIPS, AAAI, and ML4H, with notable improvements in MIMIC-III and ADNI datasets.

• Key Awards: TCS Presidential Fellowship, NSERC CGS-D Doctoral Fellowship, CMU CMLH Digital Health Innovation Fellowship, Suresh Konda Memorial Award, and multiple undergraduate honors.