Gabriel A. Terejanu
Dept. of Computer Science & Engineering
College of Engineering & Computing
University of South Carolina
315 Main St, 3A50 Swearingen
Columbia, SC 29208
Decision Making under Uncertainty
My research interests lie primarily in the field of model validation and uncertainty
quantification, information fusion, and decision making under uncertainty.
These include topics such as:
- Bayesian inference and model validation of complex physical systems, including
optimal experimental design using information theoretic concepts.
- Improving the decision making under uncertainty using accurate uncertainty
representations and goal-oriented propagation of uncertainty through computational models.
- Near-real time data assimilation with applications to evolution of toxic clouds
in air as result of chemical and biological releases.
- Applications of machine learning concepts to reduce the dimensionality of
complex uncertainty quantification problems.
- Grant: (PI) NSF-IUSE project focused on developing statistical models to track student knowledge, suggest remedial interventions, and guide future examinations in cornerstone engineering classes with high student-to-faculty ratios. Joint work with Juan Caicedo and Charles Pierce (Civil Engineering @ University of South Caroline).
- Grant: (Co-PI) NSF-DMREF project focused on linking experiments and theory in a rigorous manner using Bayesian statistics to accelerate the design and discovery of novel multimetallic catalysts for biorefinery industry. Joint work with Andreas Heyden and Salai Ammal (Chemical Engineering @ University of South Carolina) and Jesse Bond (Biomedical and Chemical Engineering @ Syracuse University).
- SPRING 2016: Teaching CSCE 590 Data Visualization,
See course syllabus.
- Our paper "Validating Predictions of Unobserved Quantities" has been accepted for
publication in Computer Methods in Applied Mechanics and Engineering.
- CONGRATULATIONS to Jiting Xu for successfully defending his M.S. thesis "Approximate Bayesian
Computation Based on Progressive Correction of Gaussian Components"
- FALL 2014: Teaching CSCE 883 Machine Learning in partnership with
EagleEye Analytics. Get the opportunity
to work on real-world insurance data and be selected for an internship interview.
See course flyer
- Our paper "Data partition methodology for validation of predictive models"
has been accepted for publication in Computers and Mathematics with Applications.