Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal J-ENSITEC (Journal of Engineering and Sustainable Technology)
Niko Akbar
University of Jambi
Civil Engineering, Building, Construction & Architecture Computer Science & IT Engineering Industrial & Manufacturing Engineering Mechanical Engineering

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

Found 1 Documents
Search

 1 
Comparison of Machine Learning Algorithms (SVM, Random Forest, and Naïve Bayes) for Predicting Rice Production Oki Dahwanu; Nurul Abdillah; Niko Akbar; Hamzah Alghifari
J-ENSITEC Vol. 12 No. 02 (2026): June 2026
Publisher : Universitas Majalengka

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31949/j-ensitec.v12i02.18386

Abstract

Global rice production faces mounting pressure from population growth and climate change, yet traditional statistical models fail to capture the complex nonlinear dynamics between environmental factors and crop yields. To address this gap, this study systematically compares the accuracy of three machine learning algorithms, Support Vector Machine (SVM), Random Forest (RF), and Naïve Bayes (NB) for predicting rice production fluctuations due to climate change using the latest local climate data from Indonesia. A dataset of 96 monthly observations (2018–2025) comprising climate features (temperature, humidity, wind speed, precipitation, cloud cover, sunshine duration) and rice production categories (Low, Medium, High) was analyzed. Algorithm performance was evaluated using accuracy, precision, recall, and F1-score. The results demonstrate that Random Forest significantly outperforms the other methods, achieving an accuracy of 95%, precision of 0.9571, recall of 0.95, and F1-score of 0.95, compared to SVM (75% accuracy) and Naïve Bayes (70% accuracy). This study provides the first head-to-head comparison of these three algorithms for rice yield prediction in Indonesia using current climate data. The key benefit over pre-existing approaches is the empirical confirmation that ensemble learning, particularly Random Forest, offers superior predictive reliability for crop yield forecasting under high feature complexity, thereby enabling more accurate, data-driven agricultural policy and food security planning.
Co-Authors Alghifari, Hamzah Nurul Abdillah Oki Dahwanu