Table 1 Root mean squared error (RMSE) of different regression methods for prediction of radon concentration in soil with respect to different window sizes.

From: Delegated Regressor, A Robust Approach for Automated Anomaly Detection in the Soil Radon Time Series Data

Win Size

Extreme Gradient Boosting (XGBoost)

Support Vector Machine Linear (SVML)

Support Vector Machine Radial (SVMR)

Delegated Regressor Method (DRM)

1

2505.088

6108.028

4826.279

1809.784

2

2473.006

6149.988

4977.869

1806

3

2416.071

6225.425

5318.782

2017.899

4

2518.619

6365.508

6527.553

1927.861

5

2593.956

6531.951

6699.957

1731.9

6

2670.358

6745.238

7109.096

2479.699

7

2761.739

6945.34

7704.349

2200.264

8

3033.962

7150.594

9223.404

1991.526

9

3135.076

7223.843

9424.501

2269.731

10

3138.239

7396.725

9659.676

2291.885

11

3596.169

7568.502

10229.27

2300.619

12

3757.667

7618.878

10881.65

2667.437