Research Article
[Retracted] Active Learning Query Strategies for Linear Regression Based on Efficient Global Optimization
Table 2
Percent improvements of the AUCs of the mean RMSEs, CCs, and Oppo Costs over BL (Ridge was the regression model; the best performances are in bold).
| | | EGO | EMCM | EGO-EMCM | QBC | EGO | GSx | EGO | RD | RD | -QBC | -GSx | -EGO |
| RMSE | Mean | 3.09 | 11.03 | 10.57 | 11.69 | 11.16 | 16.17 | 15.19 | 6.86 | 8.57 | std | 0.65 | 11.12 | 9.92 | 9.77 | 9.97 | 15.27 | 14.45 | 4.63 | 5.81 | CC | Mean | 1.81 | 4.08 | 4.40 | 5.00 | 5.01 | 7.65 | 6.41 | 4.20 | 1.96 | std | 2.35 | 3.44 | 3.42 | 3.55 | 3.40 | 2.14 | 1.58 | −1.81 | −1.88 | Oppo Cost | Mean | 56.04 | 33.75 | 45.48 | 34.58 | 44.68 | 41.97 | 51.72 | −16.38 | 50.65 | std | 34.73 | 16.99 | 24.37 | 23.07 | 26.83 | 11.68 | 20.07 | −5.19 | 31.17 |
|
|