Table 7 Results of PM and nox pollutant modeling in the grids with different Spatial resolutions (RMSE, MAE in kg/year; R2 is unitless) - (Asterisks (*) indicate the optimal values among the compared results).

From: Unravelling the importance of spatial and temporal resolutions in modeling urban air pollution using a machine learning approach

MLP

PM

NOx

RMSE

MAE

R2

RMSE

MAE

R2

Grid 500 m

19.779

8.414

0.74

*307.496

90.528

0.90

Grid 750 m

19.324

5.973

0.82

461.505

100.768

0.85

Grid 1000 m

*13.509

5.037

0.90

568.380

162.502

0.76

Back to article page