Analisis Prediktif Menggunakan Metode Hybrid Seasonal Autoregressive Integrated Moving Average – Artificial Neural Network pada Data Konsentrasi PM2.5 Harian di DKI Jakarta

I Kadek Pasek Kusuma Adi Putra (Universitas Airlangga, Indonesia)
Sediono (Universitas Airlangga, Indonesia)
M. Fariz Fadillah Mardianto (Universitas Airlangga, Indonesia)
Elly Pusporani (Universitas Airlangga, Indonesia)

G-Tech : Jurnal Teknologi Terapan
Vol 8 No 1 (2024): G-Tech, Vol. 8 No. 1 Januari 2024

Air pollution, particularly PM2.5, poses a global threat that proves challenging to address. This particulate matter has significant environmental, health, and economic repercussions in DKI Jakarta. The aim of this research is to predict PM2.5 concentrations with exceptional accuracy. The study introduces an innovative prediction model, namely Hybrid ARIMA-ANN. In comparison to ARIMA and ANN individually, this model demonstrates outstanding performance, achieving an R2 of 0.9012, MAPE of 13.603%, and RMSE of 4.061 on the training data. Evaluation on test data yields an RMSE of 5.961 and MAPE of 7.622%, indicating excellent predictive capabilities over the next 14 periods. These findings offer crucial insights for the government to forecast air quality in the future and enhance policies addressing air pollution in DKI Jakarta. The implications of these conclusions also align with the achievement of SDGs, particularly goals 3, 13, and 15.