Integrated Rearrange Processing of Hybrid Model with Weighted Values for PM2.5 Forecasting

This paper suggests using the integrated two hybrid models, ARIMA-LSTM and LSTM-ARIMA forecasting model, with weighing values for PM2.5 forecasting. The dataset, PM2.5 data of Bangkok, included wind direction (WD), wind speed (WS), temperature (Temp), and particulate matter 2.5 (PM2.5) for hourly data updated. The experiment results found that two hybrid models forecasting by ARIMA-LSTM & LSTM-ARIMA with then adjusting the weights ratio can predict the lower error when comparing forecasting with the other algorithms. Specifically, forecasting by a set of hybrid models with an LSTM Univariate Variable.