Predicting tree survival in agroforestry systems using machine learning classification algorithms

¹ Reshetnev Siberian State University of Science and Technology, 660037, Krasnoyarsk, Russia.
² Bauman Moscow State Technical University, Artificial Intelligence Technology Scientific and Education Center, 105005 Moscow, Russia
³ Krasnoyarsk State Agrarian University 660049, Krasnoyarsk, Russia

^* Corresponding author: rhfdwjdr1@gmail.com

Abstract

This article discusses the application of machine learning algorithms to predict the survival of trees in agroforestry systems. Forests play a key role in maintaining ecological balance and biodiversity, but their survival is subject to many threats, including climate change, anthropogenic impacts, diseases and pests. The study used a dataset containing data on various factors affecting the survival of trees, such as the content of phenols, the presence of arbuscular mycorrhizal fungi (AMF), lignin and non- structural carbohydrates (NSC). The classification model was built using the C4.5 decision tree algorithm, which demonstrated high accuracy (86.02%) in predicting the survival of trees. Correlation analysis revealed that phenols and AMF are the most significant factors determining the survival of trees. These results highlight the importance of biochemical and symbiotic factors for tree health. The article also discusses the importance of various factors and suggests directions for future research aimed at improving the management of forest ecosystems in agroforestry systems. The use of machine learning methods allows not only to improve the accuracy of forecasting, but also to develop more effective strategies for the conservation and sustainable management of forests.