Energy Data Science
SPE DSEATS Datathon: Oil Presence Prediction & Physics-Informed ML
Engineered machine learning and domain-aware physics-informed models for the SPE DSEATS Africa 2026 Datathon to predict subsurface hydrocarbon presence from petrophysical datasets.
Formulated a dual-track submission strategy blending Gradient Boosting with physics-constrained SVM modeling, lithology-driven KNN imputation, and physical boundary clamping.
Machine LearningPythonGeosciencePhysics-Informed MLScikit-Learn