Machine Learning Models

Here, we provide the source codes for our machine learning models and pretrained models used in the study titled "Global blue carbon losses from salt marshes exceed restoration gains." Each script is accompanied by a corresponding README file, and a sample test dataset is also included for reference.The pretrained models were used to evaluate the performance of multiple machine learning algorithms (including Random Forest, Decision Tree, and Support Vector Machine) on predicting soil organic carbon (SOC) density. This comparison helped us identify Random Forest as the most suitable algorithm, which was then used in the final SOC prediction model.