Analysis of the Use of the XGBoost Algorithm for Predicting the Behavior of Contaminants in International Aquatic Ecosystems

This paper represents a research in the field of predicting the behavior of contaminants in aquatic ecosystems. To determine the behavior of contaminants, the XGBoost machine learning algorithm was used, which is a powerful tool for analyzing environmental data due to its ability to integrate complex data sets and provide accurate predictions. The approach to predicting the behavior of contaminants can contribute to the assessment and forecasting of potential natural disasters in river-type aquatic ecosystems. This research aims to explore how the XGBoost algorithm can be used to predict the behavior of contaminants in rivers in the Republic of Moldova. A comparative analysis of the algorithm's use for rivers from various countries was carried out, as a result of which it was found that the algorithm can predict pollution levels with minimal error, providing a solid basis for the implementation of environmental policies.