Marcos Matabuena
Assistant Professor of Epidemiology
Research Interests
Professor Matabuena's teaching and research interests span digital biostatistics and epidemiology, with emphasis on functional and distributional data analysis, uncertainty quantification (e.g., conformal prediction), machine-learning-based survival modeling, and multilevel methods for continuous, high-frequency biosignals. He develops methodology for random objects in metric spaces, interpretable machine learning, and rigorous study design for complex surveys and clinical cohorts. His applied work spans diabetes-particularly continuous glucose monitoring and glucodensity profiles-physical activity and aging using accelerometry and smartphone data, and population-health analyses leveraging nationally representative datasets.
He is currently focused on advancing methods for longitudinal digital clinical data from wearables and smartphones and on integrating these streams with complementary modalities-such as genetic and other omics data-to improve prediction, inference, and decision-making in real-world healthcare settings.
Email
Prior to joining MBZUAI, Professor Matabuena was a Postdoctoral Research Fellow in Biostatistics at the Harvard T.H. Chan School of Public Health, where he worked with Professor Jukka-Pekka Onnela on statistical AI for digital biomarkers and population health. He previously spent two years in a hospital epidemiology unit, an experience that grounded his work in clinically meaningful endpoints and rigorous study design. His research introduced glucodensity as a distributional representation for continuous glucose monitoring and helped establish one of the first frameworks for uncertainty quantification with random objects in metric spaces, enabling principled prediction and inference beyond traditional Euclidean settings.
- Ph.D. in Biostatistics and Machine Learning from the University of Santiago de Compostela.
- Master of Science in Mathematical and Applied Statistics from the University of Santiago de Compostela.
- Bachelor of Science in Theoretical Mathematics from the University of Santiago de Compostela.
- Institute of Mathematical Statistics (IMS) Travel Award, 2022.
- Lugosi, G., & Matabuena, M.: "Conformal and kNN Predictive Uncertainty Quantification Algorithms in Metric Spaces." arXiv preprint arXiv:2507.15741, 2025.
- Matabuena, M., Vidal, J. C., Padilla, O. H. M., & Sejdinovic, D.: "Kernel biclustering algorithm in Hilbert spaces." Advances in Data Analysis and Classification, 1-42, 2025.
- Matabuena, M., Ghosal, R., Aguilar, J.E. et al.: "Glucodensity functional profiles outperform traditional continuous glucose monitoring metrics." Sci Rep 15, 33662 https://doi.org/10.1038/s41598-025-18119-2, 2025.
- Matabuena, M., & Crainiceanu, C. M.: "Multilevel functional distributional models with application to continuous glucose monitoring in diabetes clinical trials." arXiv preprint arXiv:2403.10514, 2025.
- Matabuena, M., Petersen, A., Vidal, J. C., & Gude, F.: "Glucodensities: A new representation of glucose profiles using distributional data analysis." Statistical methods in medical research, 30(6), 1445-1464, 2021.