Claudia Bella Hardiana
Universitas Tanjungpura

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

KLASIFIKASI SEBARAN SPASIAL HOTSPOTS DI KALIMANTAN BARAT MENGGUNAKAN METODE SUPPORT VECTOR MACHINE Naomi Nessyana Debataraja; Evi Nopita Sari; Joannes Fregis Philosovio Anugrahnu; Claudia Bella Hardiana
EPSILON: JURNAL MATEMATIKA MURNI DAN TERAPAN Vol 19, No 1 (2025)
Publisher : Mathematics Study Program, Faculty of Mathematics and Natural Sciences, Lambung Mangkurat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20527/epsilon.v19i1.14682

Abstract

Forest fires are a serious environmental issue that often leads to significant ecological and economic damage. In response, efforts have been made to predict and mitigate the impact of forest degradation by improving classification techniques. This study aims to use Support Vector machine (SVM) method to analyze the spatial distribution of forest fires in West Kalimantan. The model uses historical hotspot data as the dependent variable and considers various independent variables, including slope, topography, distances from rivers, settlements and roads, and land cover types. Due to the non-linear nature of the data, the SVM has been enhanced using the Gaussian Radial Basis Function (RBF) kernel. Optimal performance was achieved with cost (c) parameter values of 2, resulting in a classification accuracy of 75.86%. These findings suggest that, when supported by an appropriate kernel function, SVM is a viable and effective approach for mapping and analyzing forest fire-prone areas. The model provides a valuable decision-making tool for forest fire prevention and land management planning in regions with similar environmental characteristics.