Babak Shahbaba

Professor of Statistics · University of California, Irvine

My work sits at the intersection of statistics, machine learning, and the biomedical sciences: developing flexible, scalable methods in statistical machine learning and applying them to problems in neuroscience, biomedicine, and health. I integrate this research with my teaching, training students and postdocs to connect rigorous statistical reasoning with real scientific problems.

I received my PhD from the University of Toronto (2007) and did a postdoc at Stanford (2008). I am an elected Fellow of the American Statistical Association, former Director of UCI’s Data Science Initiative, and currently lead PI of an NIH–NIGMS T32 training program in biostatistics.

Statistical Foundations of AI Uncertainty Quantification Bayesian Computation Learning Across Heterogeneous Data Latent Representation Learning Neuroscience & Health

babaks@uci.edu GitHub ORCID

Latest: our NIH–NIGMS T32 training program STEER (2025–2030) supports eight PhD students per year. If you are interested, please apply through the STEER website. More news →

Focus areas

Three connected threads run through my current work.

Statistics & AI

How statistics should evolve as AI becomes central to scientific discovery and decision-making — treating shared information, generalization, and uncertainty as part of model construction, and building models that are flexible yet robust.

Learning across heterogeneity

Multimodal, optimal-transport, and graph-based methods that share information selectively — learning across people, experiments, and data sources without erasing meaningful variation.

AI for scientific discovery

Representation learning integrated with temporal and decision models, so that AI can move beyond pattern recognition toward testable accounts of how complex systems change, plan, and respond to intervention.

Liang, Z., Qu, A., & Shahbaba, B. (2026). Meta Fusion: A Unified Framework for Multimodality Fusion with Mutual Learning. Journal of the Royal Statistical Society, Series B (to appear).
Miao, R., Shahbaba, B., & Qu, A. (2025). Reinforcement Learning for Individual Optimal Policy from Heterogeneous Data. Annals of Statistics, 53(4), 1513–1534.
Yuan, Y., Zhang, Y., Shahbaba, B., Fortin, N., Cooper, K., Nie, Q., & Qu, A. (2025). Optimal Transport based Cross-Domain Integration for Heterogeneous Data. Journal of the American Statistical Association, 120(551), 1449–1462.
Moslemi, Z., Liang, Z., Fortin, N., & Shahbaba, B. (2025). Multi-Graph Meta-Transformer: An Interpretable Framework for Cross-Graph Functional Alignment in Neural Decoding. NeurIPS 2025 AI for Science (spotlight; runner-up, best poster award).
Ren, Y., Shahbaba, B., & Stark, C. (2025). Identifying dementia neuropathology using low-burden clinical data. Alzheimer’s & Dementia, 21(8), e70539.
Zhou, W., Qu, A., Cooper, K., Fortin, N., & Shahbaba, B. (2024). A Model-Agnostic Graph Neural Network (MaGNET) for Integrating Local and Global Information. Journal of the American Statistical Association, 120(550), 1225–1238.
Moslemi, Z., Meng, Y., Lan, S., & Shahbaba, B. (2024). Scaling Up Bayesian Neural Networks with Neural Networks. Transactions on Machine Learning Research.
Duan, J., Ngo, M. N., Karri, S. S., Tsoi, L. C., Gudjonsson, J. E., Shahbaba, B.*, Lowengrub, J.*, & Andersen, B.* (2024). tauFisher accurately predicts circadian time from a single sample of bulk and single-cell transcriptomic data. Nature Communications, 15, 3840.
Tran, B., Shahbaba, B., Mandt, S., & Filippone, M. (2023). Fully Bayesian Autoencoders with Latent Sparse Gaussian Processes. International Conference on Machine Learning (ICML 40).
Shahbaba, B., Li, L., Agostinelli, F., Saraf, M., Cooper, K. W., Haghverdian, D., Elias, G. A., Baldi, P., & Fortin, N. J. (2022). Hippocampal ensembles represent sequential relationships among discrete nonspatial events. Nature Communications, 13, 787.

All publications →

News