Research Article
Spatiotemporal Approaches for Quality Control and Error Correction of Atmospheric Data through Machine Learning
Table 8
Estimated results from the final normal data using SVR.
| Weather element | #data | Average | MLQC-HT | MLQC-NT | Correlation coefficient | RMSE | Time (sec) | Correlation coefficient | RMSE | Time (sec) |
| Temperature | 9346 | 8.5291 (5.9146) | 1.0000 | 0.0516 | 1376 | 1.0000 | 0.0514 | 4111 | Humidity | 6229 | 77.0694 (21.1415) | 0.9998 | 0.4256 | 1043 | 0.9998 | 0.4240 | 2522 | UV-rays | 5432 | 8.4278 (13.6254) | 0.9909 | 2.0303 | 583 | 0.9910 | 2.0243 | 1337 | PM2.5 | 9613 | 21.8624 (20.9066) | 0.9877 | 3.0947 | 1366 | 0.9877 | 3.0933 | 2709 | Solar radiation | 9896 | 0.0064 (0.0114) | 0.9176 | 0.0045 | 1244 | 0.9484 | 0.0036 | 1518 | u | 6844 | −0.6097 (0.8803) | 0.8066 | 0.3117 | 255 | 0.8101 | 0.3089 | 360 | | 6788 | −0.2789 (1.1347) | 0.8102 | 0.3751 | 308 | 0.8110 | 0.3746 | 493 | Average | — | — | 0.9304 | 0.8990 | — | 0.9354 | 0.8971 | — |
|
|
The numbers in parentheses are standard deviation. |