MANAGEMENT STRATEGIES FOR BIODIVERSITY CONSERVATION USING DATA ANALYTICS AND MACHINE LEARNING

Abstract

The focus of this research is to determine the ways and manner in which the application of machine learning (ML) algorithms may be used to improved methods of biodiversity conservation. Among these four algorithms –Independent Random Forest or RF, Support Vector Machines or SVM, K-Nearest Neighbor or KNN, and Gradient Boosting Machines or GBM – the suitability and the success rate in terms of the prediction of distribution of species and the improvement of the existing management practices in conservation were compared. Based on a large dataset, this work discovered that GBM had the biggest accuracy of 92% and F1-score of 0. 89, thus performing better than the RF model which had an accuracy of 89% and an F1 score of, 0. 85. SVM and KNN were also found to give reasonably good results with accuracies of 87% and 85% respectively. These results have revealed that GBM is outperformed other classifiers in analyzing environmental data and made accurate prediction. Collectively, the research highlights how the ML methodology could be applied with the remote sensing data to solve the problems of maintaining the biological diversity. Finally, the research employs a comparison of the stated machine learning algorithms with comparison of differences on practical use of these algorithms in conservation indicating how continued innovation in machines can enhance the usefulness of conservation efforts.