Ph.D. student, Earth & Planetary Sciences — Rutgers University–New Brunswick Machine learning · Geochemistry · Mineralogy · Geoscience data management
I apply machine learning to multivariate problems in the geological and planetary sciences. I focus on how geochemical and mineralogical data are represented (compositional data transformations, feature engineering, dimensionality reduction) and how those choices affect model performance and interpretability.
Current projects
- Magnetite trace-element chemistry for iron ore deposit-type discrimination
- Global mantle geochemistry characterized through multiple ML methods
- Unit-cell → mineral chemistry prediction for planetary XRD (CheMin / Curiosity, Gale crater)
- Zircon REE classification of metamorphic vs. igneous origin
- Petrographic AI/ML foundation models and thin-section digitization (Ward's ARC collection)
Working thesis: Multi-modal data-driven mineralogy across P-T-X conditions (Advisor: Dr. Shaunna M. Morrison)
🗄️ GeoCORK
GeoCORK is a desktop application and workflow for managing, storing, and querying U-Pb geochronology data in a SQL database. It began as my undergraduate thesis at CSU Fullerton with Dr. Kathryn Metcalf and now includes integration pathways with EarthChem and SESAR2 (IGSN sample import).
- 📄 Metcalf, K. & Burges, J.M. (2026). GeoCORK: An improved workflow for U-Pb geochronology data management. Applied Computing and Geosciences. doi:10.1016/j.acags.2026.100364
- Find GeoCORK on GitHub Here
- Presented at GSA, AGU, and Lamont-Doherty's Geoinformatics Research Group lecture series
- Burges, J.M., Yakymchuk, C., DiMaio, M.P., Lui, T.C.C., Prabhu, A., & Morrison, S.M. (2026). Metamorphic or igneous zircon? Classification through rare earth elements, machine learning, and logistic regression. Journal of Metamorphic Geology. doi:10.1111/jmg.70054
- Metcalf, K. & Burges, J.M. (2026). GeoCORK. Applied Computing and Geosciences. doi:10.1016/j.acags.2026.100364
- In prep: Mantle heterogeneity through multiple ML methods (G³); Ward's ARC petrographic thin-section digitization and data rescue (Geoscience Data Journal)
Languages: Python · R · SQL · Java · C++ Dev: Git/GitHub · JetBrains · RStudio GIS: ArcGIS Pro OS: Linux (Ubuntu/Raspbian) · Windows · macOS
B.Sc. Geological Sciences & B.Sc. Computer Science, CSU Fullerton (2024) · California GIT
