Perbandingan Algoritma Random Forest dan Decision Tree untuk Memprediksi Cuaca di Kalimantan Barat

Nur Ayu Fitri Maharani
Putri Yuli Utami
Rizki Surtiyan Surya

Jurnal Ilmiah Teknik Informatika dan Komunikasi
Vol. 5 No. 3 (2025): November: Jurnal Ilmiah Teknik Informatika dan Komunikasi

This research aims to determine the results of weather predictions in West Kalimantan. This research was conducted to compare the Random Forest and Decision Tree algorithms in predicting weather in West Kalimantan. The data used comes from the BMKG West Kalimantan Climate Station from 2022-2024 including attributes, namely, date, average temperature (Tavg), average humidity (RH_avg), rainfall (RR), duration of sunlight (ss), and average wind speed (ff-avg). The initial stage of research using the CRISP-DM method is to collect climate data obtained from the official website of the Meteorology, Climatology and Geophysics Agency (BMKG), then process the data, normalize the data, and then apply the Random Forest and decision tree algorithms using the RapidMiner tool. . Based on the results of weather prediction research with a comparison of Random Forest algorithm test data with an accuracy of 94.64% and Random Forest algorithm training data with an accuracy of 94.05%. And to compare the decision tree algorithm test data with an accuracy of 93.45% and the decision tree algorithm training data with an accuracy of 92.26%.