Gully feature mapping using machine-learning and Sentinel 2 imagery

In this project, I apply machine-learning methods to identify gully features in satellite images. I classify the images into two classes (binary classification) namely; gully and no gully. Specifically, I use Random Forest and Support Vector Machines classifiers to classify Sentinel 2 images of a study area in Bari region, Puntland, Somalia. RF perfomed better than SVM at 97.7% and 95.3% OA respectively.