SISTEM KLASIFIKASI KUALITAS UDARA DALAM RUANGAN MENGGUNAKAN ALGORITMA MACHINE LEARNING

Seyarum Melati; Agus Nur Khomarudin; Bunga Citra Lestari

Authors

Seyarum Melati Politeknik Pertanian Negri Payakumbuh
Agus Nur Khomarudin Politeknik Pertanian Negeri Payakumbuh
Bunga Citra Lestari Politeknik Pertanian Negeri Payakumbuh

Keywords:

Kualitas Udara Dalam Ruangan, Klasifikasi Machine Learning, Random Forest, ESP32

Abstract

Indoor air quality directly impacts human health and productivity, necessitating an adaptive and responsive monitoring system. This research develops an indoor air quality classification system based on edge computing utilizing machine learning algorithms. Data were collected from three locations namely office spaces, classrooms, and residential rooms for thirty days at five-minute intervals. This experiment generated a total of 25,920 data samples covering PM2.5, PM10, CO2, VOC, temperature, and humidity parameters. Following the pre-processing and feature selection stages using a combination of Pearson correlation and Recursive Feature Elimination (RFE) methods, the data dimension was reduced to five main parameters by eliminating the temperature feature. Subsequently, three classification algorithms were trained and tested, namely Random Forest, Support Vector Machine (SVM), and K-Nearest Neighbors (KNN). Air quality status was classified into three categories namely Good, Moderate, and Unhealthy. The evaluation results indicated that the Random Forest algorithm achieved the highest performance with an accuracy of 96.2 percent, a precision of 0.96, and a recall of 0.95. This optimal classification system was successfully implemented on an ESP32 microcontroller with an inference time of less than 50 milliseconds per sample. This system, integrated with the Blynk platform, is highly suitable for real-time air quality monitoring in smart building architectures

References

I. P. A. Arthawan and A. A. A. W. Putra, "Sistem IoT Berbasis Edge untuk Pemantauan Kualitas Udara Perkotaan Real-Time," Jurnal Indonesia: Manajemen Informatika dan Komunikasi, vol. 7, no. 2, pp. 282-290, 2026.

A. Jirapattarasakul, N. Jirapattarasakul, I. Thanyacharoen, P. Akkaralaertsest, and T. Yingthawornsuk, "Comparative performance of machine learning models for indoor air quality classification," Journal of Environmental Health and Technology, vol. 11, no. 1, pp. 45–56, 2025.

S. Karunakaran, J. Smith, and L. Wang, "Multi-source sensor data integration for enhanced air quality forecasting using ensemble methods," Building and Environment, vol. 248, pp. 111–123, 2024.

X. Li, et al., "An Embedded Machine Learning System Method for Air Pollution Monitoring and Control," AIMS Environmental Science, vol. 12, no. 4, pp. 26–42, 2025.

R. Mumtaz, et al., "An IoT-Based Air Quality Monitoring System and Machine Learning-Based Prediction," IEEE Access, vol. 9, pp. 45231-45248, 2021.

A. B. Prasetyo, R. Hansen, and H. Haeruddin, "Analisis komparasi algoritma machine learning untuk klasifikasi kualitas udara indoor berbasis sensor low-cost," Jurnal Teknologi Dan Sistem Komputer, vol. 11, no. 2, pp. 88–97, 2023.

C. M. Rosca, "Machine Learning Applications with Sensors for Indoor Air Quality Research," MDPI Sensors, vol. 26, no. 9, pp. 2909, 2026.

M. Sahan, et al., "Comparison of Machine Learning Algorithms (Random Forest, SVM, KNN) for Indoor Air Quality Classification," Environmental Monitoring and Assessment, vol. 196, no. 4, pp. 342, 2024.

L. A. Sutrisno, "Pengembangan Sistem Monitoring Kualitas Udara di Lingkungan Berbasis Internet of Things (IoT)," Sistemasi: Jurnal Sistem Informasi, vol. 15, no. 2, pp. 491–502, 2026.

L. T. Wong, K. W. Mui, and T. W. Tsang, "Updating Indoor Air Quality (IAQ) Assessment Screening Levels with Machine Learning Models," International Journal of Environmental Research and Public Health, vol. 19, no. 9, pp. 5724, 2022.