DOI: 10.36347/sjpms.2021.v08i10.001

Pages: 184-196

The latest trend of air pollution and variables influencing the air quality in Malaysia are studied in this research since there have been changes recently. Living conditions and health have been negatively impacted by air pollutants. An important method utilised nowadays is time series modelling, which is able to forecast events in the future. In this research, forecasting used one-year hourly Air Pollution Index (API) information originating from a station in Klang, Malaysia. The API values were predicted via the Artificial Neural Network model (ANN) and Autoregressive Integrated Moving Average model (ARIMA). Each of the approach’s performance was assessed via the root means square error (RMSE), mean square error (MSE), and mean absolute error (MAE). The outcomes highlight the fact that compared to ARIMA, the ANN provided the lowest forecasting error to predict API. As such, the ANN may be regarded as a reliable predictive method to generate data for the general public regarding the status of air quality at a particular time.