Support vector machine analysis to predict the value of tin exports as an effort to prevent environmental damage in the Bangka Belitung islands province

Abstract The Bangka Belitung Islands Province is one of the largest tin producers in Indonesia, so tin mining and environmental damage continue to be hot topics among the people of Bangka Belitung. Therefore, research is needed to predict the value of tin exports as part of efforts to prevent environmental damage. The analysis method used in this research is a Support Vector Machine (SVM). The SVM method is included in the supervised learning category with a machine learning approach in creating algorithms to predict results accurately. The data used is data on the tin export value of the Bangka Belitung Islands Province for the period January 2017 to December 2023. Based on the results of the data analysis carried out, it was found that the SVM method that is suitable for predicting export value is the Radial Basis Function (RBF) kernel. Based on this method, the highest value obtained in the Bangka Belitung tin export data was $219,028,197 in May 2025 and the lowest was $2,752,377 in June 2026. It can be concluded that the predicted results of the Bangka Belitung tin export data fluctuated slightly downward in line with the decline in the commodity value of tin in this province. The hope is that the decrease in tin export predictions must also be anticipated by local government policies in dealing with the availability and type of damage caused by this non-renewable natural resource commodity.