Table 1.
Studies on forecasting air pollutants using different models.
| Study | Research subject | Type | Method |
|---|---|---|---|
| Feng13 | SO2, NO2, CO, PM2.5, PM10 and O3 | S | RNN;RF |
| Li et al.14 | PM2.5 and NOx | S | RF; BRT; SVM; XGBoost; GAM |
| Yan et al.15 | PM2.5 | S | CNN;LSTM |
| Awan et al.16 | CO, NO, NO2, NOx, and O3 | S | LSTM |
| Dairi et al.17 | NO2, O3, SO2, and CO | H | IMDA-VAE |
| Lu et al.18 | PM2.5 | H | OR-ELM-AR |
| Du et al.19 | PM2.5 | H | 1D-CNNs; Bi-LSTM |
| Yafouz et al.20 | PM2.5, O3, SO2, NO2, CO, AQI | H | SVR-LSTM |
S: Single machine learning method, H: Hybrid machine learning model.