Digital technologies for forest monitoring in the Baikal natural territory (short paper).

The paper considers the problem of forest resources monitoring over large areas on the example of the Baikal natural territory. As the main data source, we use Sentinel-2 remote sensing data due to their regularity, broad coverage, multispectral parameters of the resulting image. The Random forest and Support Vector Machines (SVM) machine learning algorithms were used to classify land cover from the Sentinel-2 products. Both methods have shown good results with a fairly high accuracy. The training was carried out with data labeled manually into 12 classes.