Jakarta Air Quality Classification Based On Air Pollutant Standard Index Using C4.5 And Naïve Bayes Algorithms

2025-04-14T08:01:47+00:00

Increasing air pollution in DKI Jakarta has become an increasingly pressing environmental issue, which has a direct impact on public health and environmental sustainability. Therefore, it is very important to have a system that manages data-based air pollution levels. The purpose of this research is to classify air quality in DKI Jakarta through Air Pollutant Standard Index (ISPU) data. This data consists of parameters such as dust particles (PM10, PM2.5), sulfur dioxide (SO2), carbon monoxide (CO), surface ozone (O3), and nitrogen dioxide (NO2), as well as two classification algorithms used, namely C4.5 and Naïve Bayes. This research also seeks to compare the effectiveness of the two algorithms based on ISPU data collected in 15 Jakarta areas. The approach used in this research is to divide the data using three ratio scenarios, namely 70% : 30%, 80% : 20%, and 90%: 10%. In addition, performance assessment is carried out using accuracy, precision, recall and f1 score metrics. The experimental results showed better performance of C4.5, with an average accuracy of 95%, precision of 99%, recall of 94% and f1-score of 97%. In contrast, Naïve Bayes recorded an average accuracy of 81%, precision of 93%, recall of 73% and f1-score of 82%. These findings corroborate the validity of the C4. 5 algorithm is more effective in air quality classification based on ISPU, thus making it a reliable resource for air quality monitoring and management in DKI Jakarta, as well as supporting decision-making in air pollution control policies.

Analysis of Interrelationships between Weather Parameters in North Jakarta and Central Jakarta Based on Predictions Using LSTM and GRU

2025-04-13T12:32:40+00:00

This study analyzes the interrelationships between weather parameters, including average temperature (Tavg), relative humidity (RH_avg), rainfall (RR), and average wind speed (ff_avg) in North Jakarta and Central Jakarta, and compares the performance of Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) models in predicting these parameters. Data was collected from Tanjung Priok Maritime Meteorological Station in North Jakarta and Kemayoran Meteorological Station in Central Jakarta from December 2021 to December 2024. The results show that GRU performs better in North Jakarta, with RMSE of 9,02, MSE of 81,28, and MAE of 4,21 at 75 epochs, while LSTM yields RMSE of 10,02, MSE of 100,34, and MAE of 4,62 at 50 epochs. Conversely, LSTM outperforms GRU in Central Jakarta, with RMSE of 8,96, MSE of 80,22, and MAE of 4,65 at 100 epochs, while GRU produces RMSE of 9,53, MSE of 90,78, and MAE of 4,85 at 75 epochs. GRU is more effective in capturing extreme fluctuations, while LSTM excels in predicting interrelationships between parameters. This study provides insights into selecting the appropriate weather prediction model based on the priority of prediction accuracy or the ability to capture extreme fluctuations

SAGA: Journal of Technology and Information System

Jakarta Air Quality Classification Based On Air Pollutant Standard Index Using C4.5 And Naïve Bayes Algorithms

Analysis of Interrelationships between Weather Parameters in North Jakarta and Central Jakarta Based on Predictions Using LSTM and GRU