Recognition and classification of natural resource images in satellite images using the neural network algorithm

The paper describes an algorithm for recognizing and classifying natural resource images in satellite im-ages using machine learning methods with a tutor. Four different approaches were used to train the training dataset: Gaussian Mixture Model, Random Forest, Support Vector Machines, and K-Nearest Neighbors. The parameters of the neural network architecture were experimentally established, which allows classifying natural resources with maximum accuracy.