Comparison of Support Vector Machine (SVM) and Autoregressive Integrated Moving Average (ARIMA) Methods for Predicting Air Quality Using Python and KNIME

Tiara Melati Putri Wiryawanto (UIN Sunan Ampel Surabaya)
Zuyyina Hawani (UIN Sunan Ampel Surabaya)
Muhamad Attar Ramadhani (UIN Sunan Ampel Surabaya)

J Statistika: Jurnal Ilmiah Teori dan Aplikasi Statistika
Vol 16 No 1 (2023): Jurnal Ilmiah Teori dan Aplikasi Statistika

Air is the most important component for living things on earth. However, the changes that exist on earth cause problems, one of which is air pollution. Human activity is one of the causes of air pollution. This is what makes future air quality feasible to predict. To predict air quality, this research use the Support Vector Machine (SVM) and Autoregressive Integrated Moving Average (ARIMA) methods. For SVM itself, it presents one of the methods of machine learning techniques. Meanwhile, ARIMA presents one of the methods of the statistical model. Using data from the Open Data Jakarta website regarding measurements of the Air Pollution Standard Index (ISPU) at five air quality monitoring stations (SPKU) in DKI Jakarta Province in 2021, an analysis was then carried out to compare the performance and accuracy of these two methods in predicting air quality. The results of this study indicate that between ARIMA and SVM testing, it can be said that SVM testing has higher accuracy results. This can be seen from the average accuracy results with several treatments.