Research

Research Interests

Spatial statistics
Bayesian methods
Hierarchical models
Graphical models
Latent factor models
Environmental health

Much of the data in environmental health is spatial and multivariate, with complex dependence and missingness, which I often approach through the lenses of Bayesian methods and dependence learning. I build interpretable, scalable models; advance identifiable and robust inference for complex environmental systems; and translate results into open tools and actionable evidence for communities.

Publications, Preprints, and Work in progress

^† Equal contribution.

Browse the same publication list by research area or chronologically.

2026 Jin, B., Peruzzi, M., and Dunson, D., Bag of DAGs: Inferring directional dependence in spatiotemporal processes. Bayesian Analysis, 21(1): 401-422. [Publisher] [Package] [Code]
2024 Jin, B., Herring, A.H., and Dunson, D., Spatial predictions on physically constrained domains: Applications to Arctic sea salinity data. Annals of Applied Statistics, 18(2): 1596-1617. [Publisher] [Package] [Code]
In prep Jin, B., Finley, A.O., and Datta, A., Bayesian spanning tree-based multivariate spatial model. [Slides]
2023 Jin, B., Dunson, D., Rager, J.E., Reif, D., Engel, S.M., and Herring, A.H., Bayesian matrix completion for hypothesis testing. Journal of the Royal Statistical Science: Series C, 72(2):254-270. [Publisher] [Code]
2025 Jin, B., and Datta, A., Identification and consistent estimation in source apportionment using geometry. [arXiv] [Code]
In prep Jin, B., Hagan, D., and Datta, A., GeomNMF: Identifiable non-negative matrix factorization in Python for environmental source apportionment.
2025 Jin, B., Agrawal, P., Sato, K., Lemson, G., Hao, L., DeCarlo, P., Koehler, K., and Datta, A., Baltimore Air Quality Dashboard. [Dashboard]
2025 Jin, B., Salmerón, B.D., McClosky, D., Hagan, D., Dickerson, R.R., Spada, N.J., Deanes, L.N., Aubourg, M.A., Ramprasad, T., Schmidt, L.E., Sawtell, G.G., Heaney, C.D., and Datta, A., Source apportionment of air pollution burden using geometric non-negative matrix factorization and high-throughput multi-pollutant air sensor data in Curtis Bay, Baltimore, USA. [arXiv] [Code]
In prep Jin, B., Eyni, A., Waugh D., and Datta, A., Filling the meteorological desert: Calibrated low-cost sensors map neighborhood heat in Baltimore.
In prep Jin, B.^†, Aubourg, M.A.^†, Salmerón, B.D., Hagan, D., Dickerson, R.R., Spada, N.J., Ramprasad, T., Livi, K., Chen, J., Schmidt, L.E., Sanchez-Gonzalez, C., Sawtell, G.G., Datta, A., and Heaney, C.D., Practical multi-methods approach to characterize coal dust and air pollution burden in communities bordering coal export terminals.