Comparative analysis shows XGBoost's superior accuracy for wax deposition in petroleum fluids, indicating the value of machine learning in flow assurance.
Key Points
XGBoost achieves superior predictive accuracy for wax deposition, showing R2 values up to 0.957.
Key thermodynamic descriptors, such as the weighted average carbon number, significantly influence wax deposition behavior.
The study incorporates uncertainty quantification and interpretability diagnostics, enhancing model reliability in diverse conditions.
Hybrid physics-informed approaches provide opportunities for improved integration of domain knowledge in predictive modeling.