This repository contains the supporting files for the article:
Active Learning for Transport Property Prediction in CO₂–Hydrocarbon Systems: A Multi-Fidelity Approach Integrating Molecular Dynamics and Experiments
This repository is divided into two main folders. Each main folder is organized into subdirectories corresponding to the binary and ternary systems studied: CO₂/n-heptane, CO₂/benzene, toluene/n-hexane, and CO₂/ethanol/dibenzyl ether.
-
Active_Learning/Contains scripts and datasets used in the active learning workflows:- Experimental and MD datasets
- Python scripts for Multi-fidelity Gaussian Process (integrating MD and experimental data)
- Python scripts for Single-fidelity Gaussian Process (MD-only or experimental-only)
-
MD_Simulations/Contains input and analysis files for molecular dynamics simulations:- Force field topology and parameter files
- LAMMPS simulation scripts
- Python scripts for calculating mutual diffusivity and viscosity