Penggunaan Metode Random Forest, Support Vector Machine dan Artificial Neural Networks dalam Prediksi Suhu Udara di Balikpapan

Erika Meinofelia (Universitas Negeri Semarang Badan Meteorologi Klimatologi dan Geofisika)
Mochamad Aryono Adhi (Universitas Negeri Semarang)
Achmad Fahruddin Rais (Pusat Riset Limnologi dan Sumber Daya Air BRIN)
Djunaidi Djunaidi (Universitas Negeri Semarang)
Feddy Setio Pribadi (Universitas Negeri Semarang)

BIOEDUSAINS: Jurnal Pendidikan Biologi dan Sains
Vol. 8 No. 6 (2025): BIOEDUSAINS:Jurnal Pendidikan Biologi dan Sains

This study aimed to compare the performance of three algorithmic models, namely Random Forest, Support Vector Machine (SVM), and Artificial Neural Networks (ANN), in predicting air temperature in Balikpapan. Changes in air temperature influenced by various climatic and geographical factors present a major challenge in urban planning; thus, accurate predictions are crucial to support sustainable and climate-adaptive city planning. The dataset used consists of observational data from the Balikpapan Meteorological Station, BMKG, over ten years, from January 2014 to December 2024. The models were evaluated using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and R Squared (R²) metrics. The results show that the SVM method produced an MAE of 0.17, RMSE of 0.21, and R² of 0.95, providing better predictions than ANN and Random Forest. In conclusion, SVM is an effective method for air temperature prediction in Balikpapan. Keywords: Artificial Neural Networks, Random Forest, Support Vector Machine, Machine Learning, Air Temperature Prediction