Density-Based Spatial Clustering of Applications with Noise (DBSCAN) and Principal Component Analysis (PCA) for Anomaly Detection

Hanna Arini Parhusip (Master of Data Science Departement, FSM /UKSW)
Suryasatriya Trihandaru (Master of Data Science FSM UKSW)
Bambang Susanto (Master of Data Science FSM UKSW)
Adrianus Herry Heriadi (PT Artha APuncak Semesta Indonesia (APSI))
Petrus Priyo Santosa
Yohanes Sardjono (BRIN Jogjakarta)
Johanes Dian Kurniawan (Master of Data Science FSM UKSW)

Lontar Komputer: Jurnal Ilmiah Teknologi Informasi
Vol 15 No 02 (2024): Vol. 15, No. 2 August 2024

This research addresses a critical issue in industrial environments: air quality, specifically regarding PM 1.0 and PM 2.5. High concentrations of these particles pose significant health risks. The study measures temperature, humidity, pressure, altitude, PM 1.0, and PM 2.5 and shows the effectiveness of using AIOT-Particle devices to analyze these features with Principal Component Analysis (PCA). Density-Based Spatial Clustering of Applications with Noise (DBSCAN) is used to detect anomalies during the observation period. Anomalies occur when the altitude ranges from 65 to 70 units, according to PM 1.0 and PM 2.5 values. The positions where anomalies occur are illustrated based on altitude, temperature, pressure, and concentration. The results demonstrate that altitude dominates as the first feature. Finally, the research concludes that altitude, PM 1.0, and PM 2.5 are the dominant features. The study confirms the effectiveness of PCA and recommends using these three features for anomaly detection in DBSCAN. Overall, the research highlights the novelty and success of AIOT-Particle in industrial environments.