Air Quality Prediction in DKI Jakarta Using Support Vector Machine: A Comprehensive Classification Approach

Purpose: This study develops a machine learning-based classification model to predict Air Pollution Standard Index (ISPU/AQI) categories in DKI Jakarta using the Support Vector Machine (SVM) algorithm. It addresses the challenge of accurate multi-class air quality classification under highly variable urban pollution conditions and imbalanced class distributions.
Research Methodology: A quantitative approach was employed using 1,825 daily observations from the Satu Data Jakarta portal collected between February and November 2023. Six pollutant variables (PM2.5, PM10, CO, SO₂, NO₂, and O₃) were used as predictors. Data preprocessing included missing value imputation, duplicate removal, label encoding, and Min-Max normalization. An SVM model with a Radial Basis Function (RBF) kernel was implemented in Python (Scikit-learn) on Google Colaboratory. Hyperparameters were optimized using GridSearchCV with StratifiedKFold cross-validation. Model performance was evaluated using accuracy, precision, recall, F1-score, and confusion matrix analysis.
Results: The optimized SVM-RBF model achieved 96.1% classification accuracy on the test set. Macro-average F1-scores reached 86% for the Good category, 98% for Moderate, and 93% for Unhealthy. Performance for Very Unhealthy and Hazardous categories was lower due to class imbalance and limited observations.
Conclusions: The proposed model provides accurate and reliable AQI classification for urban tropical environments and offers a practical foundation for automated real-time air quality monitoring systems.
Limitations: The study is limited to DKI Jakarta and a ten-month observation period, while minority classes remain underrepresented.
Contributions: This study presents a validated SVM-based AQI classification framework that can support scalable, data-driven air quality monitoring and environmental decision-making in Indonesian metropolitan areas.