Scholar

Gabriel A. Terejanu

Google Scholar ID: Z7LP12kAAAAJ

Professor, University of North Carolina at Charlotte

Causal AIInvariant Machine LearningBayesian InferenceUncertainty Quantification

Citations & Impact

All-time

Citations

1,161

H-index

i10-index

Publications

Co-authors

list available

Contact

Publications

20 items

Browse publications on Google Scholar (top-right) ↗

Resume (English only)

Academic Achievements

Published 'Robust Machine Learning via Gradient Consistency' (arXiv:2411.06040)
Published 'Concept-Driven NOTEARS (CD-NOTEARS)' at ICMLA 2023 (DOI:10.1109/ICMLA58977.2023.00118)
Published work on invariant molecular representations in Journal of Chemical Information and Modeling (DOI:10.1021/acs.jcim.3c00594)
Published prostate cancer disparities study in Annals of Surgical Oncology (DOI:10.1245/s10434-024-15675-1)
Led or contributed to projects funded by NSF, ARO, USDA/NIFA, Lowe’s Innovation Fund, and Toyota Racing Development

Research Experience

Prostate cancer disparities analysis with Atrium Health, investigating the impact of socioeconomic status (SES) on high-risk diagnosis
Invariant NASCAR tire modeling project with Toyota Racing Development, using invariance principles to align tire models with experimental and track data
Developing a causal AI engine for Lowe’s to generate actionable operational recommendations
Leading aflatoxin prediction project funded by NIFA/USDA to build a general predictive framework for mycotoxin incidence in crops
Causal modeling of affective polarization funded by ARO, integrating data-driven discovery with expert knowledge
Deep learning for surface chemistry funded by NSF, combining computational catalysis and neural networks to predict chemical behavior
Proposed CGLearn, a robust ML method based on gradient consistency across datasets
Developed CD-NOTEARS, extending NOTEARS with concept-level causal priors
Researched invariant molecular representations using Siamese networks for accurate adsorption energy prediction

Co-authors

25 total